Milos Maricic has a unique view into how the world’s largest asset owners are using AI.

Maricic and his team at SPEC Research work with sovereign wealth and pension funds as they figure out where the technology actually helps with investing.

Over the past six months, his team has also tried to answer a straightforward question: What are the world’s largest sovereign wealth and pension funds actually doing with AI?

They reviewed more than 1,000 public sources, incorporated information supplied by institutions and contacted the funds directly to assess 27 sovereign wealth funds and 26 pension funds. And today they released their SWF AI Maturity Index report.

The big takeaway:

The world’s largest sovereign wealth and pension funds are rapidly adopting AI, but none can show that it has made them money.

Or as CPP Investments CEO John Graham summed up in the report: “Will [AI] help us make faster decisions? Yes. Will it help us make better decisions? TBD.”

What Maricic and his team found was a wide gap between adoption and proof.

Singapore’s GIC has built a Virtual Investment Committee that includes an AI devil’s advocate trained on 44 years of internal deal data. The agent argues against an investment before the real committee sees it. CPP Investments has issued more than 2,100 Microsoft Copilot licenses across a firm of roughly 2,100 people, with 73% actively using them.

The strongest claim comes from Norway’s sovereign wealth fund, which says AI saves it about $100 million a year in trading costs. The fund has not published a methodology attributing those savings to the technology. More broadly, most funds do not track which parts of an investment decision came from AI and which came from people, making it difficult to tell whether AI improved a decision or simply made the process faster.

I spoke with Maricic about how sovereign wealth and pension funds are using AI. We talked about what sovereign wealth and pension funds are actually doing with the technology, why managers’ claims of AI-generated returns often fall apart under scrutiny, and why the growing volume of AI-generated research is creating another problem: the cost of verifying it.

I turned this interview, with the help of AI, into a new episode of my Alpha Intelligence Podcast. It has been dormant, but I plan to share more interviews going forward in this format, alongside the written transcripts.

Matt: How long have you been putting this report together?

Milos: About six months now. It is the first thing of its kind, so it has been ramping up really fast. I am just looking at how the response from the institutions has changed. When we launched the first one six months ago, it was like, “Do you want to comment on this?” And they were like, “No. Who are you? Just get out of here.”

Now it is more like, “Ah, yeah, sure. Let’s have you talk to our AI team. Let’s calibrate that.” So it has been getting some pickup.

ICYMI

Matt: So you are reaching out to these funds directly, saying, “Hey, we are trying to get a better sense of how you are thinking about this technology?”

Milos: Yeah. There are three things. First of all, the baseline job is to go through everything that is publicly available. The index has a database of more than 1,000 public sources.

Some of these institutions have internal information about themselves or their clients, and they allow us to use it specifically. Sometimes they do not, and that is fine. The third layer is that we reach out to them and say, “Hey, we are going public with this in a couple of weeks. Please help us understand what is going on.” More often than not, they come back, so we feel like we have a relatively robust picture of where the use is. Is it perfect? No. But this is a super-fast-moving field.

Matt: Right. And there are not many other resources looking at it like that.

Milos: That is the thing. Sovereigns are a niche to begin with, and even those who tackle that niche do not necessarily talk about AI use within it. So it is a niche within a niche.

Matt: What made you start it?

Milos: My last company was an AI-driven capital-allocation platform that got acquired by a sovereign.

So a lot of sovereigns and pensions were clients in the latter stages of that business. Business with sovereigns is a who-you-know business. It is very relationship-driven. We felt it was interesting for the team to continue in the same direction. We also felt we were doubling down on the advantage of being a little bit of insiders, knowing the right people, and so on.

Matt: I think of your role as somewhere between the vendors and all these solutions. If you are working at a sovereign wealth fund or you are a money manager, all this AI stuff can be too much.

So you are helping them sort out what is actually working, right?

Milos: Exactly. If it is overwhelming for you and me just trying to follow what is going on, it is even more overwhelming for them because they have all that capability. But what do you do with the capability?

Even if you have bought the story that AI is going to be big and you want to be part of it, what do you do? Where do you deploy? Do you deploy in the biggest, most inflated-valuation companies out there? Do you look for the nuggets? Do you go through managers?

If we focus on the manager side, a lot of what allocators do is manager selection. There is an explosion of managers coming to them and saying, “We use AI. Look at our backtests. Look at our track record. It is all amazing.” Allocators are prudent about it, and they have every reason to be, because a lot of this is just bogus.

Some of it is interesting, but you have to distinguish between the two. One of the things we produce is a framework called the SPEC test that we coach allocators to use with managers. It breaks down their AI claims in a structured way to understand what holds water and what does not.

One priority for the second half of this year is working with the Institute for Sovereign Investors to take the SPEC framework and turn it into a more industry-wide due-diligence framework. It is sorely needed. Right now, what is happening within allocators around AI claims is completely ad hoc.

Matt: What are fund managers saying they are doing with AI?

Milos: The most typical thing is: “We backtested this, it worked amazingly well, and our Sharpe ratio ended up being three.” Then you look into the backtest and realize the model had a little glimpse of the future. It actually saw where the market was going.

Another thing is riding heavily on the track record of the managers themselves. But there is loads of research showing that, if you use AI to do fully automated investing, the aspects that make up a manager’s track record are what AI can replicate most easily. The more idiosyncratic parts are more difficult to pick up.

When we make managers go through the framework, we ask them to decompose their claims. We probe them. We say: “Tell me about a time when the model was wrong. What happened? How did you adapt it? How did you fix it going forward?” Or: “Tell me about a time when it was overruled by a human.” That gives you a window into their internal process, so you can understand where the boundary is between human capacity, human oversight, and AI. The framework has about 14 questions that we drill into.

Matt: A sovereign wealth fund will call you up and say, “XYZ reached out. I want you to take a look. Do you know about them? Maybe you already do.”

Milos: Yeah, exactly.

Matt: What are some red flags you have seen?

Read more

Alejandro Lopez-Lira has been running one of the largest public experiments in AI investing. Nearly 52,000 investors have allocated about $200 million on Autopilot across seven public portfolios built with AI models including ChatGPT, Claude, DeepSeek and Grok.

These portfolios aren’t responding to a simple “What stocks should I buy?” prompt. Instead, they break investing into a series of steps—macro analysis, company research, portfolio construction and verification—with different AI agents handling each stage.

So far, that approach has produced stronger results than many public AI investing experiments. The DeepSeek portfolio, for example, has gained 56% over the past year, compared with a 16% gain for the S&P 500. Lopez-Lira cautions that it can take a decade to know whether a diversified strategy is outperforming the market.

I recently caught up with Lopez-Lira, an associate professor of finance at the University of Florida, to see how his thinking on AI and investing has evolved since he became AI Street's first interviewee two years ago. Back then, Lopez-Lira argued that AI could already handle many of the tasks typically assigned to an intern. Today, he believes AI systems have reached roughly the level of a fourth-year Ph.D. student.

We discussed what's changed over the past two years, how he builds AI systems that break investing into specialized tasks, and why he believes the next wave of AI on Wall Street will be driven by systems rather than standalone models.

This interview has been edited for clarity and length.

How capable have AI systems become over the past two years?

I would say large language model systems are at the level of a fourth-year Ph.D. student in every field.

I don’t think there is any specific task that does not require human-to-human interaction or doing things in the physical world that can generally be done better by humans than by AI.

Jobs are collections of tasks, sometimes well-defined and sometimes not. That determines how much substitutability there is between humans and AI systems.

I say AI systems because LLMs by themselves haven’t changed that much. But embedded in systems such as Claude Code, it’s insane what you can do. I’ve been mostly working on autonomous systems that can run for a long time without human intervention.

ICYMI

The Claude Portfolio

One example is your Claude portfolio. How does it differ from the ChatGPT, DeepSeek and Grok portfolios?

The other models receive a list of the best potential stocks from a large scoring system. Grok decides how to make its portfolio, GPT decides how to make its portfolio and so on.

Claude is more iterative. It can launch an agent for each stock to check that the research is sound and that there are no issues. You can also have a portfolio-optimization agent assemble the portfolio and another agent audit it to make sure there’s nothing weird.

It involves many more cycles and revisions, calls more agents and potentially trades more frequently. The other portfolios trade once a month. Claude decides when to trade, with a limit of twice per week. It trades less often than that on average.

How has it performed?

It was launched about four months ago. It was up about 10% as of yesterday, roughly the same as the S&P 500.

Four months is way too little time to tell. It will take a couple more years to evaluate.

The ChatGPT portfolio has outperformed the S&P 500, but it trails Grok and DeepSeek. What explains the difference?

I think ChatGPT is more opinionated. DeepSeek says, “These seem like the stocks with the highest expected returns. I’m just going to make a bet on that.” That seems to work.

ChatGPT says, “I’m going to consider what’s happening with the macro environment and other things.” That may not necessarily be correct. Grok is somewhere in between.

The system produces very good information, and the models use it more or less effectively. But it’s still too early to draw conclusions about any of them.

How much money is invested across the portfolios?

There is $200 million across all portfolios now.

Some experiments that let LLMs trade have produced large losses and volatile results. Why haven’t your portfolios experienced the same swings?

Read more

BlackRock began using Aladdin, its in-house portfolio- and risk-management system, in the late 1980s and started selling it to clients in the 1990s. Originally built for institutional investors, Aladdin has since expanded into a broader platform spanning wealth management, trading and private markets, with BlackRock acquiring eFront and Preqin along the way.

Aladdin now has more than 130,000 users and accounts for most of BlackRock’s $2 billion technology-services and subscription business.

More recently, BlackRock has started adding AI to its products and investment process. An Aladdin Wealth feature used by Morgan Stanley draws on portfolio holdings, client preferences, risk data and the bank’s market outlook to draft commentary for financial advisers. Internally, BlackRock’s Asimov system uses AI agents to scan research notes, company filings and emails for information that could affect portfolios managed by its fundamental-equities teams.

BlackRock is also rolling out RockAI, an internal platform that allows employees to create specialized agents without writing code.

To understand how BlackRock is approaching the next phase of that work, I recently spoke with Dhagash Mehta, one of the researchers behind its experiments with AI agents. I’ve previously covered his research here and AI developments at BlackRock:

Mehta runs applied AI for investment management at RockAI. He spent the previous three years in Aladdin Financial Engineering, the firm’s centralized quant group, working on trading, liquidity and portfolio optimization. His job is to take problems from portfolio managers and traders and turn them into models that could eventually run on Aladdin — with the access controls, guardrails and audit trail required to use them.

This interview has been edited for clarity and length.

What makes building an AI platform for a regulated industry different?

We have to have many guardrails, and access management is a big thing because we always have to make sure that certain data is not accessible to certain people. And tracing as well—tracing the whole thing and making everything auditable. Plus, there are different developers who all might have different rights.

How do you select the research topics?

Previously, I was in academia for many years—some would argue too long. There, whenever I used to choose a problem, it was basically from my head. I would create a toy model, create a problem out of thin air, and try to solve it. Then I switched to industrial research, and I am realizing that real-world problems are even more challenging and even more interesting.

My strategy is—I am not running a completely academic AI lab here. My eventual goal is still to deploy models for us, and they might have some very good use for our portfolio managers and traders, eventually on Aladdin. The problems here come from the actual PMs and traders. Most of the time, they start talking about some problems in their language, and then I translate it to the machine learning and AI language. Most of the time, no off-the-shelf method would solve this problem as is. I always have to innovate something, even some small thing. Then I realize that maybe no one else has thought about this in the literature, so that is my opportunity to write a paper.

How do those problems usually reach you?

Because I am embedded in the business, we always have frequent meetings with PMs and traders. We also have planning for the next year, the next quarter, and so on. That is when we go through many discussions, and that is where all these problems come up. PMs might say, “This is our burning problem right now. Can you help solve this for us?” That is how it starts.

Your agent paper used three agents to assess companies. How did that work?

Read more

If you’re a '90s kid like me, you remember buying CDs at Circuit City. You couldn’t buy individual tracks online, so you had to buy the whole album for the one or two songs you wanted.

Technology eventually unbundled music. Now it is beginning to do the same to financial data.

Traditionally, an equities trader who needed data for only a handful of stocks might still have to license an exchange’s entire market-data feed.

This problem was the genesis of Databento, which lets customers buy specific market data and pay based on usage.

Christina Qi co-founded the Utah-based company after seeing the challenges of acquiring market data while running a high-frequency trading hedge fund. Founded in 2019, Databento announced last week that it raised a $97 million Series B led by NEA with participation from DRW Venture Capital, Redpoint Ventures and Tribe Capital.

Qi has leveraged her HFT background at Databento by placing the company’s servers in the same data centers as exchanges to receive their market feeds directly. The company processes market data into standardized formats and sells real-time and historical data for equities, futures and options through APIs. With the new funding, the company plans to expand to more than 20 data centers worldwide and has secured more than 100 petabytes of storage capacity to support that growth.

I recently spoke with Christina about how technology is disaggregating market data, a shift I’ve been covering (see my interview with Carbon Arc CEO Kirk McKeown and Aiera COO Gavin Skinner). Qi and I discuss fundraising in an AI-obsessed VC environment, why AI labs are buying market data, the complexity of exchange licensing and how usage-based pricing can create problems for large firms.

This interview has been edited and condensed for clarity.

Seems like venture capital investors are focused on backing AI companies right now. What was it like raising in this environment?

Christina: We are actually one of the few non-AI companies to raise a round of this size, which is quite unusual in the current market environment. While many non-AI companies face pressure to position themselves as AI businesses, we have never felt that pressure. We got this far by being hyper-focused on what we are good at, which is data. We choose not to compete against our customers, who excel at AI. Instead, we stick to our strengths to support them.

Market data has typically been sold in bundles. Why do you sell it a la carte?

Christina: We are the first company to bring what is called usage-based pricing to this industry, meaning instead of paying for the entire exchange—like buying the entire grocery store—you can browse items, taste samples, and buy a bento box for dinner. This allows individuals and teams the freedom and flexibility to shop around before committing to larger datasets.

It is also product-led growth in the sense that customers can self-service and sign up on their own. They do not need to talk to a salesperson. All the pricing is online: here is the price, here is what you get, register, and check out.

More AI Street Expert Interviews

Companies like Carbon Arc offer alternative data on an a la carte basis. Do you see other companies in the data space moving toward this model, and is this the future of the industry?

Christina: There are many incredible startups in our space executing different applications. But it’s worth mentioning that a la carte pricing only scales to a certain extent. Large customers, such as banks and consulting firms, have a fixed budget for data, which works better with annual, flat-rate pricing. Usage-based pricing works great for early-stage experimentation, but can cause administrative headaches down the road.

You mentioned previously that major AI labs are buying data from Databento. What are their primary use cases?

Read more

I thought the technology had a real shot at becoming widespread across Wall Street. Or, if not, it was going to be a great big dud worth covering.

Early on, I spoke to Snowflake’s Jonathan Regenstein, who was also optimistic that AI would have a big impact on how the investing world works. I interviewed him in the fall of 2024, and much of that conversation focused on a basic question: how do you even use this technology for investing? Does it make sense to? What about hallucinations?

When we spoke again recently, Regenstein said the skepticism has largely disappeared, at least among the most sophisticated funds and asset managers. The question now is how to build efficient processes.

Regenstein is a former attorney and self-taught coder who now leads wealth and asset management at Snowflake, where he works with banks, hedge funds and asset managers building AI systems on top of their own data. Snowflake does not build frontier models. It works with providers and platforms including OpenAI, Anthropic, Mistral, Meta and Hugging Face, which gives Regenstein a view into how firms are putting different models into production and where they still struggle.

He says investors are using AI in three main ways:

1. Chat: using the model conversationally to test a hypothesis or get a first read on what’s happening.

2. Text analytics: “Go examine this 10-K for me” — or, at asset-manager scale, turning years of filings, research papers, and sell-side research into an “engine of insights.”

3. Coding assistant: “Write the code for me” — helping quants and investors code up hypotheses, reproduce papers, build dashboards, or call validated models.

Regenstein has also published a new book, Large Language Models: The Hard Parts: Open Source AI Solutions for Common Pitfalls, coauthored with Thársis Souza PhD, whom I’ve previously interviewed on my podcast here. The book is for non-experts who want to understand LLMs well enough to use them: builders, go-to-market salespeople and strategy leaders. I’m looking forward to getting my copy in the mail!

This interview has been edited for clarity and length.

Matt: Tell me about your role.

Jonathan: I work at Snowflake, which is a super interesting place to work right now because we don’t build frontier models. We partner with OpenAI, Anthropic, Mistral, Meta and Hugging Face. We partner with all of them.

As this world evolves, we’re sitting at the center of it, even though we’re not really as high up in all the headlines. We see it all, deal with it all, work with them all and optimize for all of them. I get to see some trends emerging around who uses which tool for what.

There are three different ways that LLMs are being deployed now. One is the chatbot. People like the chatbot. The second is text analytics: “Go examine this 10-K for me.” The third is a coding assistant: “Write the code for me.”

Those are three very different things. The models can do all of them, but they’re not the same thing at all. They’re not for the same person or the same use case. They’re really totally different. It’s starting to break down along those lines, and there’s a lot underneath each one.

How has AI adoption in finance evolved since we first spoke in the fall of 2024?

Jonathan: There have been a few big shifts. Back in 2024, I think you and I were both believers in this. But at that time, with some exceptions, if you had gone into a room full of quants or portfolio managers, I think you would have gotten a lot of crossed arms: “I’ve been doing this a long time. I understand, and I’m sure AI is helpful, but it’s not going to revolutionize how I do my job.”

We’ve now come to a place where there is no more skepticism, certainly at the most sophisticated shops and asset managers. The question is: “How can we optimize this?”

It’s becoming a recruiting and retention tool for funds and asset managers: “You should come work here as a quant. Why? The AI harness we’re going to give you is better than you’re going to get anywhere else. If you don’t have that, you cannot win in this world.”

Once you’ve come to a fund or asset manager and they’ve built you that harness, you don’t want to leave because you can’t take it with you. It’s intellectual property. It’s becoming a recruiting and retention tool. It’s a complete shift.

How are firms building governance into those AI harnesses and testing whether they actually work?

Jonathan: This is an area where I work directly. Snowflake doesn’t build its own models, so what are we doing? We’re trying to give you the governance architecture that sits underneath them.

When you ask for data, we need to make sure that what you asked for is what you actually got and that it is underpinning your answers. That’s a ton of the work we’re doing right now: validating your context and making it scalable to monitor and evaluate what you’re doing.

Skills are becoming a huge thing. What skills do is bring determinism, as much as possible, to the way large language models work because you’re giving them a recipe: “Do this. Follow these steps. Bang, bang, bang, bang, bang.”

One of those recipes can be a quantitative model that I’ve built somewhere. Large language models don’t run a stress test or even run a regression for you. They could write the code to run the regression, but I don’t want the model to write the code on the fly. I want it to call the model I’ve already built and validated.

Let’s say I have a phenomenal modeling team and they’ve built these models. We’ve validated them and have them in our ecosystem. This is something you can do with Snowflake: You can store all your models in one place.

Now the LLM needs to choose the right model, not run the model. Where we have a whole model-testing framework that we’ve validated over the last 20 years, the LLM can call the right model. That’s what I need to get right more than I need to get an LLM to build a model on the fly for me.

You can do that for research. Part of the research for your model garden can be having an LLM build a model for you. But when it’s actually time to make a decision, the LLM needs to call a validated, back-tested model.

A lot of the focus is on the model, but much of the work involves the surrounding environment and calling the right document or model. How do you ensure the system calls the right thing? Do you use knowledge graphs?

Read more

The Federal Reserve’s interest rate decision is among the most well-known, market-moving government release. Each statement is instantly parsed by newswires, redline tools, and analysts looking for tiny shifts in language.

Smaller, less-liquid markets have historically lacked that kind of infrastructure. Why invest heavily in technology to trade small markets, like orange juice futures? Trading ends up choppy and concentrated among a few dominant players.

Advances in AI may help narrow that gap by lowering the cost of processing market-moving information and building custom software around smaller, more specialized datasets. One firm trying to build around that idea is Moreton Capital Partners, a systematic commodities investment manager founded by Les Finemore and Alistair Fullerton.

The two were previously at Farrer, a commodities hedge fund that launched in 2024 with backing from Millennium Management. The investment firm redeemed after a second-quarter 2025 drawdown in Farrer’s fundamental portfolio, ending an exclusivity arrangement, according to Finemore. He said the redemption allowed him and Fullerton to launch Moreton for other clients.

The firm has raised about $100 million through separately managed accounts, with plans to top $500 million by year-end and reach $1 billion by next spring, according to a representative. Trading in the SMA accounts began a few weeks ago, so performance data is limited. Moreton has also been producing trade ideas for a multi-strategy hedge fund.

MORE AI STREET INTERVIEWS

I spoke with Finemore about why he thinks AI can give commodities investors an edge. He pointed to the USDA’s monthly World Agricultural Supply and Demand Estimates report, known as WASDE. Released in the middle of the trading day, the roughly 40-page report includes detailed forecasts for crops including oilseeds, cotton, sugar and poultry —too dense a document for a human to digest quickly.

Finemore said Moreton’s system spotted an opportunity in wheat after the May WASDE report.

In our conversation, we discuss:

• How Moreton is trying to systematize fundamental commodities trading across 100+ markets.

• Why Finemore says commodity trading still relies on manual work, physical-market knowledge and fragmented data.

• How the firm uses LLMs to test trading ideas, write code, find replacement datasets and interpret model outputs.

• Why Moreton built its own risk platform, and how it wants the system to explain portfolio moves through Slack and a web interface.

• Why Finemore thinks portfolio managers at multi-strategy funds may not have the time or incentive to build unproven AI infrastructure.

This interview has been edited for length and clarity.

Matt: When you say you are using AI to invest in commodities, what does that mean exactly?

Les: We are forecasting more than 100 commodity markets. We trade metals, energy, agricultural commodities and soft commodities, including markets in China through swaps.

Most of our trading happens on a weekly basis, when we rebalance the book. We are market neutral. We look more like a long-short equity fund than a concentrated commodity manager.

We think about it as systematizing the fundamental trader. Traditional quants often focus on price, volume and momentum. We use a lot of fundamental data: weather, physical-market prices, basis differentials at the interior and export levels, landed cost-and-freight prices, satellite data and shipping data. We also use options, flow and macro data.

We have more than 250,000 datasets with many features. A human trader may only be able to cover five markets closely in a day. We are trying to look across many markets while retaining that fundamental detail.

Matt: What does AI let you do in commodity markets that you couldn’t do before?

Les: May’s WASDE report was our first proof of concept trading the release this way. We fed in market-survey data, positioning and the price drift before the report. Then wheat production came in with a big surprise.

We could quickly determine: ‘This is a big miss on production. This is bullish.” Wheat then went limit up. There was still time to position.

That is where we see an edge: finding inefficiencies in smaller markets that not everyone is watching and analyzing the release inside a window that would previously have been too short.

At a previous firm, a trader would print the WASDE report, underline it and perhaps trade 45 minutes later. That workflow is too slow now.

Matt: There’s been a lot of growth in systematic strategies in stocks. What’s the trend in commodities?

Les: Commodity trading is still very manual. I started exporting grain from Western Australia and later traded at Merricks Capital, where a four-person investment team covered more than 30 markets.

We traded fundamental dislocations across commodities and calendar spreads. The fund performed relatively well, but the process was not very repeatable. A lot of the work could be automated: gather the information, narrow it down for the trader and systematize how the balance sheet is deployed.

When you cover 30 markets, it is hard to know each one intimately and consistently find an edge. I came away thinking there had to be a better process than pointing, clicking and entering data into Excel.

Matt: How do you use LLMs?

Les: We are trying to use them at each step of a traditional quant’s workflow.

We did natural-language processing in 2017 and 2018, but it was a very naive approach. Now, LLMs make it much easier to take unstructured data and factor it into models quickly.

Every day, we scrape academic articles and blogs looking for potential signals. If an academic is writing about an idea, it is probably already saturated, but it can still give us something to build on. An LLM can help write the code needed to test the hypothesis and build a model around it.

Agents can also search for potential data sources. They can identify a dataset that has been deprecated and suggest a replacement, or find a new dataset that looks different from what we already have.

Then there is interpretability. At Imbue, we had thousands of model outputs and struggled with how to interpret them. We now use LLMs over those results.

Everything is trying to augment the human.

Matt: Are the LLMs producing trading signals?

Read more

Dixon, who co-wrote the textbook Machine Learning in Finance, worked in structured credit, and holds a PhD in applied math from Imperial College London, argues that talent, models, and mathematical elegance are increasingly constrained by the hardware budget behind them.

Dixon’s point is that brute force gets you a long way in finance. Most applications are not precision engineering problems — they are approximations where more compute buys a better answer. Firms that have the resources to scale compute are pulling ahead of peers that cannot.

Compute catches up

One of the most basic questions a large bank has to answer is also one of the hardest: What is the profit and loss on my positions today?

For years, financial firms didn’t have the computing power to analyze their positions quickly. For a large bank, getting one simulation number out of tens of thousands of positions requires teraflops of processing. Stress testing and marginal analysis stack on top of that. “The bulk of the load sits in any trading strategy that involves derivatives: hedging positions, marking to market, getting an overall risk profile,” says Dixon, who now runs his own consultancy, Quiota. GPUs helped — they were already good at parallelizing simulation threads before deep learning arrived. When deep learning did arrive, it fit the GPU architecture almost perfectly.

Stale overnight

Risk models have traditionally run overnight. A market shock hits and every parameter is useless. "You wake up and there's a big news event, say, the Suez Canal is blocked, there's war with Iran, and the data regime is completely different, and all your parameters are stale." AI surrogates, deep learning models trained on the output of those expensive simulations, can now recalibrate in seconds. Instead of rerunning the whole model, you feed in the latest data and out pops a value. Dixon says tier-one banks have been pursuing this with varying degrees of success.

Compute as cost

More compute solves the speed problem but creates a spending problem. “One bank told me the cost of running jobs on rented cloud compute had become a major expense.” Significant enough to drive infrastructure decisions. And security adds a constraint — these firms don’t want sensitive positions sitting on someone else’s servers. Operational cost is now the barrier to entry, the same way an expensive platform was always the price of admission in high-frequency trading.

Research compressed

The research cycle is collapsing. “If you want a model for a new financial instrument, describe it with a few equations, and AI will generate the mathematics to figure it out. What would have taken weeks takes minutes now.” That changes the economics of building new products entirely. Engineers Dixon works with tell him they haven’t actually coded since last September — Copilot or Claude does everything. The frustration is in the last mile. “You’re on the putting green and it pulls out a driver,” he says. You get to the prototype fast, but one incremental change request and the codebase shifts so much that no one is sure what it does anymore.

Dixon was named one of Risk magazine’s 2022 buy-side quants of the year, with Igor Halperin, for machine-learning research in wealth management. This interview has been edited for clarity and length.

Matt / AI Street: What do banks and trading firms spend most of their computing power on?

Matthew Dixon: I’d break it down broadly, across buy-side and sell-side. You’ve got portfolio optimization, asset management, which is always resource-intensive. But I think the bulk of the load sits in any trading strategy that involves derivatives: hedging positions, marking to market, getting an overall risk profile. That’s where you get into expected shortfall and value at risk. Banks have to do that as a legal requirement. Those computations are where the heavy lifting is.

That was all before AI, and all of it is simulation. GPUs came along and were very good at parallelizing independent simulation threads. The problem was they had limited onboard memory, and it took a lot of latency to move data into that memory. The old saying was: you don’t get out of bed for GPUs unless you’ve got heavy compute and not too much data movement. What worked really well were risk simulation, pricing, optimization, signal detection: all compute-intensive work.

Then deep learning arrived and fit the GPU architecture almost perfectly. The hard work of mapping routines onto GPUs was already done: NVIDIA’s custom libraries, Intel’s equivalents. And suddenly you had a new possibility for derivatives pricing and risk. Instead of running complex models overnight and missing a recalibration window when the market moves, you could train an AI to learn the metasurface of a risk model. Then it becomes a lookup function. Feed in the latest data, out pops a value.

I know tier-one banks have been looking at doing this for some time, with varying degrees of success. I was approached by one, I’m not at liberty to say who, and they were explicit about what they needed. It’s really around derivative products where all the headache is.

Read more

INTERVIEW

Andy Legg has recruited for Citadel, Point72, AQR Capital Management and Two Sigma, helping drive Wall Street’s shift toward quant strategies and PhD-led teams.

Now, he’s hiring for the next phase: AI.

In a recent interview, Andy, now a Director at Riviera Partners, discusses how hiring became “PhD-centric” after the financial crisis, how the high cost of computing power is limiting who can build AI systems, and how at least one hedge fund is letting those systems make trading decisions.

PhD Hiring Took Hold After the Financial Crisis

“I started in quant recruiting in 2009, just after the crash. The majority of recruiting I’ve done ever since has been PhD-centric, and it has entirely changed the landscape of Wall Street.”

A Hedge Fund Using AI as Trader

“The most advanced AI in trading capability I’m aware of is a particular hedge fund where the AI is the trader. The researchers and engineers are feeding the reasoning engine of this AI daily to make better trading decisions, but the AI is making the trading decisions.”

Private Equity Is Running Into a Data Problem

“They [Private Equity firms] are realizing the quality of their data in terms of validation, alternative data, and structured versus unstructured data is not where it needs to be to run AI, let alone prediction or recommendation systems.”

The Seven-Figure Talent War

“Candidates in the R&D spectrum who have a CS/Stats/ML/AI PhD under a renowned professor, published relevant papers at leading conferences... can command north of seven figure total comp packages at a pretty young age.

The “Vibe Coding” Interview

“An increasing number of funds are changing their approach to acknowledge those that can use prompt engineering or vibe coding effectively. The onus is increasing now on asking what AI tools you use and how AI-savvy you are.”

This interview has been edited for clarity and length.

Matt: I feel like the image of Wall Street is still the “Masters of the Universe” type, but it’s become way more PhD-driven now.

Andy: It’s funny when you talk to people removed from finance; they think of the New York Stock Exchange and people being there Eddie Murphy-style, like in Trading Places or Wall Street, thinking everyone is still wearing suits and ties. It’s anything but that these days. Everything is algorithm-driven and automated. The difference now is the compute… very few AI startups-mid size tech firms can afford to train their own models because they haven’t got the money to pay NVIDIA for the compute.

I started in quant recruiting in 2009, just after the crash. The majority of recruiting I’ve done ever since has been PhD-centric, and it has entirely changed the landscape of Wall Street.

Matt: It seems like it’s going to be more and more that way. The OTC markets have basically been the barrier, but now if you can structure all this data, it changes things.

Andy: This evolution is no longer limited to the upper echelons of finance. We are seeing that now in private equity. The data drive we witnessed in quant maybe ten years ago—where big data providers started popping up—is now increasingly happening in PE because they want to do AI to benefit from operational efficiency. They are realizing the quality of their data in terms of validation, alternative data, and structured versus unstructured data is not where it needs to be to run AI, let alone prediction or recommendation systems.

Matt: AI has been around for a while, but LLMs have made it “AI” in the way people talk about it now. When did it start to shift recruiting-wise?

Andy: Machine Learning has been around for more than 40 years. The difference now is the compute and the scale at which it can be performed.

I placed my first machine learning research engineer at a hedge fund in 2013. If you look at some of the more illustrious quant hedge funds— Renaissance Technologies, D.E. Shaw, TGS —they were leveraging ML in the late 90s and early 2000s. G-Research is another where their ML quant group was started in the early 2010s and seeking to incorporate ML into their trading strategies.

If you watch the show Billions, which is primarily based on Point72, they do a good job of parodying elements of the market adoption and certain innovations. One of the more renowned data science to market intelligence stories of the 2010’s is about a lumber strike in Canada. The fund in question flew drones and used satellite imagery to figure out when the strike was going to end and how much lumber was piling up. These insights informed their trading strategies and its rumored they made millions of dollars in profit from this in a matter of days. Now in the 2020s, most quant funds are combining the insights data science has provided them with AI, whether gen-AI and/or agentic AI to find greater opportunities to beat the markets.

The most advanced AI in trading capability I’m aware of is a particular hedge fund where the AI is the trader. The researchers and engineers are feeding the reasoning engine of this AI daily to make better trading decisions, but the AI is making the trading decisions. This business is doing very well thus far. News broke recently that Instacart co-founder Apoorva Mehta is launching a new hedge fund, Abundance, where AI agents will work as the portfolio managers.

Matt: So, it’s truly autonomous trading?

Read more

One pattern shows up again and again: many investors tried AI a year ago or so, had a bad experience, and wrote it off because it got facts wrong, misstated figures, or invented citations.

Yet they keep hearing about AI in finance. That creates a different kind of pressure: not wanting to be left behind, without a clear sense of what has changed.

In this interview, he explains why product overload is slowing adoption, why “skills files” matter more than model training, and how to structure workflows so outputs can be trusted.

He also points to areas like fixed income, where these approaches may matter more than people expect.

The conversation also covers something that doesn’t get enough attention in finance: how bias shows up in ways that aren’t obvious. Not just demographic bias. Positional bias (the same information, presented in a different order, can produce different outputs), framing effects (the same odds stated two ways lead to different decisions), and the fact that models reflect the biases in the data they’re trained on.

We cover:

• Why many investors are still anchored to early AI failures

• Why comparing models is the wrong approach

• How “skill files” and workflows actually drive results

• Where early ROI is showing up (including fixed income)

• How bias shows up in model outputs

This interview has been edited for clarity and length.

Matt: You talk to a lot of investment professionals just getting started with AI. What’s your sense of adoption among investors?

There’s a large group that tried ChatGPT when it first went mainstream, had a bad experience — it made something up, got a calculation wrong — and wrote it off. That’s a reasonable response based on what it was at the time. The problem is anchoring. They’ve frozen their view of the technology at that moment, and the tools are genuinely different now. I tell them: forget everything you knew about this 18 months ago. You have to be experimenting again.

The other group has FOMO, but the anxiety from not knowing where to start is actually keeping them from doing anything. My advice to both groups is the same: treat it like a new employee. Give it a task. Check the output. See what it can do.

From Models to Workflows

Matt: We are seeing a total product overload right now. It isn’t like comparing phones based on pixel counts; it is very difficult to compare these AI models side-by-side. How should people navigate this?

Brian: It is very hard to compare them, and getting all of them at once can be overwhelming. I recommend trying to understand what you can do with the “Frontier” models and Claude’s skills—as well as the skills OpenAI is developing—and what can be achieved with connectors. For example, Notion already acts as an agnostic transcript writer that can connect to Claude. Many investment professionals are not yet aware of the tools that are used ubiquitously in the computer science realm.

Matt: I’ve had Claude skills on my radar for a few months, but I’m still trying to get my arms around them. How are you using them currently?

Read more

Cornell’s new AI in Finance certificate began in this vacuum. Victoria Averbukh, Professor of Practice and Director of Cornell Financial Engineering Manhattan, spent two years talking to portfolio managers, traders, and strategists before designing it.

“People would say it was about not having a clear way to think about the systems, what the system is doing,” Averbukh said. “New terminology kept coming in.”

Manage Email Preferences

If you prefer to receive one weekly email with all AI Street content, turn off Research and Interviews here:

Manage How Often You Receive AI Street

The program — 30 sessions, 13 instructors — mixes Cornell faculty with practitioners from fintech, asset management, investment banking, and trading. The goal is judgment, not tool proficiency.

Kathryn Zhao, Head of Institutional API Product, OKX, says it reflects a shift already underway in hiring. Domain experience used to be the deciding factor. Now she screens for AI awareness.

“If someone understands how to work effectively with AI tools [...] they can onboard quickly and begin contributing almost immediately,” Zhao said.

In the conversation below, we discuss:

• Why applying AI in finance can’t be a direct translation from tech

• How the certificate balances academic foundations with practitioner insight

• What AI awareness means for hiring and talent development

• The biggest stumbling blocks for AI adoption in financial services

• Why chasing the pace of change is less useful than building understanding

The below conversation has been edited for clarity and length.

Matt: What was the genesis of this certificate? When did you decide to do this, and what was the catalyst?

Victoria: I kept hearing, across different finance sectors, that people were either using AI and finding it useful but not fully comfortable with whether they could trust it, or they didn’t even know where to start. That hesitation was very consistent—for portfolio managers, traders, execution people, strategists, research people, more quantitative people, less quantitative people. It didn’t really matter. People would say it was about not having a clear way to think about the systems, what the system is doing. New terminology kept coming in. We started with AI, then the word “agent” appeared. It just felt either overwhelming or there was a lack of trust.

My light bulb went on around 2024, about two years ago. I remember that Kathryn and I actually went to have coffee at Breads Bakery on the Upper East Side, and I said, “Kathryn, I have this thought.” And Kathryn said, “Yes!” She was one of my very early supporters. That coffee at Breads Bakery is what gave me confidence to go and investigate more.

Matt: What makes applying AI in finance different from applying it in tech?

Victoria: After speaking with Kathryn, I also spoke with Andrew Chin, Marcos López de Prado, and others. They were all very clear that education is needed, partly because of the hesitation we just talked about, but also because finance is not tech, and applying AI here requires respecting that difference.

Machine learning, big data technologies, and large language models were all built for something else, not for finance. Uber’s business model, for example, is built around offering a service powered by new technology. That is fundamentally different from what a bank or a hedge fund does. So applying AI to investing, to execution, to alpha generation, or even to forecasting market exposure cannot be a direct translation. The objectives are different, and the data is different. Financial data is non-stationary, often smaller, and rarely clean, so you cannot just take machine learning methods from tech and apply them directly.

Our industry is and will continue to adopt AI, but it has to be done carefully, with a real understanding of what works and what does not. Everyone I spoke with strongly supported the idea that training is needed specifically because of these differences, and that developing critical understanding, judgment, and a clear sense of potential ROI before adoption is essential.

Which is why the real question is not whether we use AI, but how we use it in a way that actually improves decision-making rather than just adding complexity.

Matt: Can you talk about the structure of the certificate and the role of practitioners in it?

Victoria: The full certificate is about 30 sessions with 13 instructors. The curriculum is deliberately structured to start from fundamentals — faculty from Johnson School and Engineering explain what the data is, what an LLM is, and work through use cases.

But because it’s so fast-changing, you really need practitioners to understand what needs to be done. Finance is an extremely regulated industry. I think that’s another thing that differentiates it. Even probably from healthcare.

The industry instructors are very carefully curated to give breadth of coverage — fintech, asset management, investment banking, and trading. This is not a certificate just for trading or fraud detection or financial advising. It’s for everybody. Ideally you have some experience on Wall Street, but also if you’re just starting out, it’s really for everybody.

Do you know the quote from Einstein? “If you can’t explain it simply, you don’t understand it well enough.” That was my guiding principle. I know that our Cornell faculty can take the complicated topics — transformers, LLMs, all of that — and make it intuitive. Developing that intuition is really the intention behind the certificate. It’s what enables you to make sound judgments about when, where, and how AI should be used, and when it shouldn’t.

Recent Interviews

Matt: Kathryn, where do you see this heading? How do you see AI impacting finance in the next couple of years?

Kathryn: Speaking from a practitioner’s perspective, my approach to hiring has fundamentally changed. A few years ago, I would evaluate candidates primarily based on their prior experience in the specific role or industry. Today, that is no longer the deciding factor, particularly for junior hires.

What I prioritize now is AI proficiency and AI awareness. If someone understands how to work effectively with AI tools (how to ask the right questions, interpret outputs critically, and apply insights to real business problems) they can onboard quickly and begin contributing almost immediately. With access to AI-generated materials and the ability to leverage AI as a day-to-day copilot, the learning curve is dramatically compressed.

In that sense, traditional domain experience is no longer a strict prerequisite. What matters more is a strong baseline understanding of the real world at a college-educated level, combined with the ability to operate fluently in an AI-enabled environment.

That is why I believe an AI in Finance certificate program is highly relevant. It prepares participants to become AI-aware and AI-capable without requiring them to be programmers. More importantly, the AI literacy and applied mindset the program builds will open a wide range of opportunities for participants in the years ahead.

Victoria: When you say the person needs to be AI-aware—does that mean the person can have zero finance knowledge, or do you mean they don’t need deep knowledge of Python and machine learning?

Kathryn: They don’t need to come in with deep expertise in Python or extensive financial industry knowledge. Those capabilities can be developed on the job. What matters most as a baseline is their ability to work effectively alongside tools like Claude: knowing how to frame the right questions, extract the right information, and translate insights into action.

Victoria: I agree with Kathryn, but maybe a notch below the enthusiasm. Here’s why: This certificate is not about the tools. It’s about understanding the lay of the land and developing intuition. Learning what Claude does can be done on YouTube. There are plenty of tutorials.

The AI awareness Kathryn mentioned, that’s what we bring in the certificate. Ideally, as an educator, I want participants to leave thinking: I know what questions to ask. I know how to bring judgment to that Claude-generated code. So maybe we’re fast-tracking people a little bit through the first nine months on the job once Kathryn hires them.

I also think finance is segmented. You can be an expert in energy, or equities, or fixed income, or mortgages. You can be a really great financial advisor, but you wouldn’t necessarily know how to construct a global allocation as a portfolio manager. At some point, applications of AI are going to become more tailored to all these different areas. It’s almost like you’re not going to go to a dentist if you need new glasses.

Ideally, if this certificate is successful and we offer it again and again, I certainly want to make sure that we have significant participation from practitioners, from industry. Engineers will be inventing new AI 2.0 and 3.0 and 10.5, but the industry participation will always be needed. Maybe we break it up or reshape it to focus on specific areas of finance, that’s also a possibility.

Matt: What is the biggest stumbling block right now for AI adoption in finance?

Victoria: I think it’s uncertainty. I think it’s leadership that is probably older and did not grow up with phones in their hands. There’s a certain inertia. Bridging the generational gap is going to be harder. I think CEOs are going to get younger.

Matt: How do people keep up? It feels like the terminology alone is a moving target.

Victoria: There’s no glossary out there. That glossary changes dynamically. That’s going to be part of the certificate. Once people finish, they’re going to know the terms and will be more comfortable and ready for a new iteration of terms. But ultimately, I think trying to chase the pace is impossible. Focus on understanding, not the hype.

His advice: don’t start with technology. Identify work that can be automated.

Many companies are doing the opposite— buying tools first and figuring out where they fit later.

“We’re letting the vendor marketplace drive our strategy as opposed to asking the question: what do you want?”

McMillan, who recently launched McMillanAI, where he advises executives on AI strategy, says many organizations are still early in figuring out how to deploy AI at scale.

In practice, that means starting with tasks that take up a lot of time and are repeated across large teams—call centers, onboarding, compliance review. These are areas where AI can replace or augment work in a measurable way.

What breaks at scale isn’t the model. It’s everything around it: how data is structured, who has access, how systems are monitored, and how much autonomy they’re given.

Most firms haven’t solved that yet. They’re experimenting with tools, but haven’t redesigned how work actually gets done.

We cover:

• Identifying high-volume work AI can replace

• What breaks when you try to deploy AI at scale

• Why most firms are still stuck in pilot mode

• How to think about vendors vs building in-house

• Where agents are actually being used today (and where they aren’t)

This interview has been edited for clarity and length.

Matt: Why start McMillanAI now?

Jeff: There’s an enormous gap in the marketplace around education and awareness. The people that need to make the decisions — and by the way, I’ve probably spoken to no less than 30 CEOs in the last six weeks, CEOs of Fortune 500 companies — they want to do AI. They’re getting pressured to do AI. But we have a workforce that knows more about this technology than most senior people do in organizations. That’s a gap, and that’s an opportunity.

I don’t want to sound Pollyannaish about this because I’m not: this is a once-in-a-generation type of technology, and I really do believe that we have a choice as humanity. We have a choice about how we deploy this for the benefit of all of us as opposed to maybe a few. I’d like to be part of that dialogue.

Matt: Going back to those 30 conversations, what were the common threads?

Jeff: There’s an enormous amount of external pressure on them. There’s no CEO I talked to that says, “I don’t believe in AI.” That was maybe true three years ago — “Is this just the next crypto? Is it the next metaverse?” No one believes that now. Everyone believes there’s something going on here. So that’s number one.

Number two, there’s a tremendous desire to do something, but they don’t have the skills and the competencies to do this technology at an enterprise level. If you look at every major technical transformation, it takes eight to 10 years to fully play out. So, we’re very early in the process.

The problem with AI is it requires a different approach, and it’s not a technology problem.

Read more

He expects AI to have an internet-scale impact on markets:

𝘐 𝘢𝘮 𝘢 𝘣𝘦𝘭𝘪𝘦𝘷𝘦𝘳 𝘵𝘩𝘢𝘵 𝘵𝘩𝘪𝘴 𝘪𝘴 𝘪𝘯 𝘴𝘰𝘮𝘦 𝘸𝘢𝘺𝘴 𝘭𝘪𝘬𝘦 𝘵𝘩𝘦 𝘧𝘪𝘳𝘴𝘵 𝘥𝘰𝘵-𝘤𝘰𝘮 𝘣𝘰𝘰𝘮. 𝘚𝘶𝘳𝘦, 𝘪𝘵 𝘸𝘪𝘭𝘭 𝘤𝘩𝘢𝘯𝘨𝘦 𝘫𝘰𝘣𝘴 𝘢𝘯𝘥 𝘵𝘩𝘦𝘳𝘦 𝘸𝘪𝘭𝘭 𝘣𝘦 𝘫𝘰𝘣 𝘭𝘰𝘴𝘴𝘦𝘴, 𝘸𝘩𝘪𝘤𝘩 𝘯𝘰𝘣𝘰𝘥𝘺 𝘦𝘷𝘦𝘳 𝘸𝘢𝘯𝘵𝘴. 𝘉𝘶𝘵 𝘪𝘯 𝘵𝘩𝘦 𝘭𝘰𝘯𝘨 𝘳𝘶𝘯, 𝘵𝘩𝘪𝘴 𝘪𝘴 𝘢 𝘵𝘰𝘰𝘭 𝘢𝘯𝘥 𝘢𝘯 𝘢𝘪𝘥 𝘵𝘰 𝘩𝘦𝘭𝘱 𝘱𝘦𝘰𝘱𝘭𝘦 𝘥𝘰 𝘵𝘩𝘦𝘪𝘳 𝘫𝘰𝘣𝘴 𝘣𝘦𝘵𝘵𝘦𝘳 𝘢𝘯𝘥 𝘵𝘰 𝘤𝘳𝘦𝘢𝘵𝘦 𝘯𝘦𝘸 𝘫𝘰𝘣𝘴 𝘸𝘦 𝘥𝘰𝘯’𝘵 𝘬𝘯𝘰𝘸 𝘢𝘣𝘰𝘶𝘵 𝘺𝘦𝘵. 𝘐 𝘳𝘦𝘢𝘭𝘭𝘺 𝘵𝘳𝘶𝘭𝘺 𝘧𝘦𝘦𝘭 𝘭𝘪𝘬𝘦 𝘵𝘩𝘢𝘵’𝘴 𝘸𝘩𝘦𝘳𝘦 𝘵𝘩𝘪𝘴 𝘪𝘴 𝘨𝘰𝘪𝘯𝘨.

He leads market structure and technology research at Crisil Coalition Greenwich, where he tracks how banks, asset managers, and trading firms deploy new systems. He previously worked at BlackRock and TABB Group.

But AI adoption on trading desks has yet to scale.

A recent report from the firm shows the most common use cases among bond traders are still data analysis and document review, not execution or decision-making.

Trading desks face clear regulatory and reputational risk, where firms need to explain and defend decisions to clients and regulators. As McPartland puts it, you can’t tell regulators: “Well, the AI did it.”

In this interview, McPartland explains where AI is being deployed today, what’s holding back trading applications, and why coding and developer productivity may be the most important near-term use case.

This interview has been edited for clarity and length.

Manage Email Preferences

If you’d like to receive fewer emails, you can manage your subscription and turn specific sections on or off.

Manage How Often You Receive AI Street

Why AI Adoption in Trading Is Moving Slowly

Matt: I was checking your reports on AI, and you guys are focusing on how it’s in the back office. That seems like it’s the story across the street. When do you think it goes beyond that to more trading applications?

Kevin: I think we’re starting to get there. I was at FIA Boca earlier this week and there was definitely a lot of talk about AI. I think the industry is excited and interested but really trying to be cautious. There’s a reputational risk issue, a regulatory issue. You don’t want to do the wrong thing for your clients from the sell-side perspective. If there’s an issue and regulators come to you and ask what happened, you can’t just say, “Well, the AI did it, I’m not sure.” That’s not a good answer. So I think that’s leaving people cautious.

We actually just got back a study of bond traders and we asked them where they saw the opportunity in AI. Not surprisingly, data analysis was number one, document review number two. So it still really is about pouring through data and unstructured data to help digest it, find insights, find patterns. I think that’s still the biggest use case now.

My two cents — I think where really a lot of the impact will be in the short, medium, and long term is on the coding side. Everything from making the most sophisticated quant developers even more efficient than they already are, to letting business users prototype what they want in a way they never could before, and then handing it off to IT. I just think the possibilities are absolutely huge in that regard.

ICYMI

AI Regulation and Governance Are Still Catching Up

Matt: What’s your sense of what needs to happen on the regulatory and standardization side? It’s such a new technology — there are no best practices yet.

Kevin: It does need to happen, although it’s hard to put a finger on it, because almost by definition it’s not something that is structured. You could ask different LLMs or even the same LLM the same question and it might give you a different answer. So by definition, it’s not structured. But yes, maybe it’s just continuing to learn what the potential risks and pitfalls are. How do you look out for them? How do you catch them? How do you prevent them?

Of course the models themselves will continue to get better, which should limit some of those things, but it could create new ones as well. Just saying “no, it’s not safe, we can’t use it” — that’s not the answer either. This is here to stay. It’s going to have a big impact on the market. All that work is required, and I think we’re already starting to see more working groups and roundtables and people working through it, talking to their peers, trying to understand what are the best practices. What are you doing? What are you doing? So we can all sort of try to figure out the most effective way forward, because there is just a lot of opportunity.

Coding May Be the Most Underestimated AI Use Case

Matt: What do you think is underappreciated or not talked about enough in this space?

Kevin: The coding agents are talked about broadly — Claude Code and OpenClaw, that’s all over the news. But for capital markets and trading specifically, I don’t feel like it’s talked about very much. What does that look like? How is it used on a trading desk? Is it used on a trading desk yet? Are there rules there?

Matt: I spoke to Man Group. They’ve developed something called AlphaGPT. I think the hedge funds have a little more flexibility — they’re regulated, but they’re not tens of thousands of people usually. Some of the quants are doing this, but I think the technology has moved faster than the humans in terms of how they can actually put this out in a responsible way.

Kevin: Yeah. To me it all feels inevitable. It’s just figuring out how to test it and how to do it safely.

AI’s Impact on Finance May Look Like the Dot-Com Era

Matt: Is that opinion shared broadly? A year or two ago, AI in finance was not really considered as impactful as some other areas. Is the industry stance now that this is going to have a big impact?

Kevin: This industry doesn’t ever all agree on anything. I am a believer that this is in some ways like the first dot-com boom. Sure, it will change jobs and there will be job losses, which nobody ever wants. But in the long run, this is a tool and an aid to help people do their jobs better and to create new jobs we don’t know about yet. I really truly feel like that’s where this is going. Not about large-scale job loss, but people in the seats being able to do things they never knew how to, never had time for, or just never could before.

Matt: High frequency trading has gotten so fast that it’s approaching the speed of light, so you can’t really top that. You have to find other ways to make money.

Kevin: That’s right. It’s not just about speed. In equities, maybe it is, but that’s why there’s only a few dominant firms left doing it at scale. Somebody said to us a year or two ago, it’s not about being faster anymore — it’s about being smarter.

The CIO and founder of IDX Advisors says AI has helped cut legal bills, replace outsourced developers, and automate proprietary workflows over the three years since ChatGPT launched in November 2022.

His team comes from a quant and coding background, which made it easier to start experimenting.

The firm, a systematic asset manager focused on risk-managed digital asset strategies, began testing large language models shortly after ChatGPT’s release. One of the first use cases they built was a way for AI to read PDFs, something models couldn’t do three years ago.

What started as a tool for reviewing documents has now grown into a broader internal system for coding, compliance and CRM automation. The firm now runs multiple models on the same task and has them critique each other’s output before a human reviews the results. The same approach has allowed the team to replace an offshore development group and build internal tools that would previously have required outside vendors.

I’d like to think I’m pretty current with the new AI tools trying them myself, but Ben is ahead of me with OpenClaw, which he describes this way:

In our chat, Ben explains how the firm structures these AI workflows, the tools it relies on, and where he sees the biggest opportunities for AI in financial services.

He walks through how IDX built an AI-powered paralegal workflow, replaced an offshore development team with coding models, and created internal agents that automatically research and enrich potential clients.

He also explains why he believes persistent systems like OpenClaw could become a core layer of AI infrastructure inside small firms.

One theme that comes up repeatedly is that AI handles much of the grunt work while Ben and his team review and validate the results.

This interview has been edited for clarity and length.

Matt: How did you get started with AI?

Ben: I’ll give you a quick overview from day one of everything we did that was material. I’ll caveat it by saying that myself and the other founders come from a quant hedge fund background. We were coming into this already with some software development capability. We were doing our own APIs and things like that. We had machine learning models predicting Bitcoin prices. There was at least a modicum of technical experience in-house.

When ChatGPT first came out, like everybody else, we thought it was an interesting chatbot that could write poetry or create raps. But instantaneously we started just throwing things at it. A lot of people forget—it wasn’t that long ago—but the original ChatGPT couldn’t read PDFs. So the very first thing we built was a simple PDF reader. That was something we had experience with, because you have to vectorize the data. There’s OCR and all that. We did that specifically for legal. Compliance is expensive, and we’re a small business—a 10-person team with revenue under $3M, which is not low, but we need to save money where we can. Using ChatGPT to basically run our own paralegal department instantaneously cut our legal bills. I did the math: we easily saved a million dollars in legal bills since the launch of LLMs, which is material.

Matt: What were you doing previously, and how did you implement this?

Ben: Previously we had different lawyers for different things: a compliance lawyer, corporate attorneys, and JV lawyers. Everything was a back-and-forth. These are expensive Wall Street lawyers. A perfect example is new LP docs. That should be pretty “control C, control V”—a lot of that is templated. Why are we paying $75,000 for another set of LP docs?

I’ll zoom out and make a meta comment. Yes, AI is going to be disruptive—this is the new industrial revolution. But it’s also going to be hugely democratizing for small businesses. It has been tough to compete, irrespective of industry, as a small business in America for really the last 10 years. This is going to disintermediate massively in favor of small businesses.

Legal is a perfect example. We had a busy year in 2024, and what we did (regarding using LLMs in-house)—we always use a red team, blue team approach. That is the quickest way to dramatically increase the quality of the LLM output.

By giving two different LLMs the same task and have them review each other’s work. Especially when it comes to things like legal. There was that headline early 2023 about a lawyer using ChatGPT to draft a brief in which the LLM massively hallucinated, and we were cognizant of that. So we would have Claude and ChatGPT both review a document, come up with comments, and then have them review each other’s comments. We would take that to our lawyer and say, “This is our comprehensive review.” At that point, they didn’t necessarily know we were using ChatGPT, but I’m sure they were looking at it and saying, “I can’t overcharge for this.” What would have been a 40-hour exercise is now literally a two-hour exercise. We spent $7 in AI compute.

We are basically replacing their paralegal function but not paying for it. We even talked to one group that asked if we could set up a custom LLM in-house for them. People are already seeing the writing on the wall.

Matt: How did this transition into your software development?

Ben: We’re originally a quant fund, so we’ve been developing our own software for internal use for years. For things like Python or SQL database software, we’re experts. Where we were paying for heavy dev work was on anything on the front end. We wanted to create business intelligence dashboards so the whole firm could see our machine learning Bitcoin model outputs—not just me and the research team that can run Python on our computers. The problem is, when you get into front-end UIs—TypeScript, React—that might as well be hieroglyphs to us.

In Q1 2023, we had a full offshore outsource dev team—one of these software teams offshore—and we were spending easily up to five figures a month on these guys. They were good. They built us internal dashboards, took a lot of our Python scripts, and turned them into real software we used internally.

I started popping things into ChatGPT. I would prompt it and say, “You are a Chief Technology Officer supporting me, the CEO of a quantitative hedge fund.” Those long, specific prompts really help. It could take Python code and help with the front end. It would say, “Go to Vercel, spin this up, go to GitHub,” and we would have a UI push.

Fast forward through different iterations—Gemini, Claude Code—and that same offshore team eventually called us asking what the next project was. I told them we had taken it in-house. They asked how, and I said Claude Code. We run red team, blue team, so we’re running Codex and Claude Code simultaneously and having them check each other’s work. There was silence on the other end of the line.

What you come to realize is that in software, the expensive part is yes, the engineers, but there is also the cost of time. What’s nice about having embedded LLM functionality—on the legal side or the code side—is the feedback loop is virtually instantaneous. The software development cycle rapidly accelerates and it’s cheaper. I didn’t have to go back and forth on Slack or explain the logic to these guys. The LLM understands that perfectly, especially the latest versions, because they’ve got very high-functioning reasoning. LLMs are expert-level translators that speak every language on the planet, including legal and code. Up until now we’ve had to pay a lot of money for human translators in those domains, and that has just been ripped away.

Matt: You mentioned automating your lead enrichment and CRM. How did that work?

Ben: We have a lean, mean sales team. We’re quants, so we’re very big on data enrichment and digital outreach. Everything has to be run through compliance. We were looking at Salesforce and thinking about how to automate lead QA. We don’t necessarily want a junior person doing that because it’s a waste of their time—and it’s not simple QA. What we want is an agent that can look at the firm that clicked on our email, go to their website and find out who they are, then go to their SEC ADV filing—which is a public filing showing their lines of business and what type of advisor they are. We would also have the agent look at the website’s “About Us” section for anything related to golf or sailing to help enrich the conversation. We wanted all of this to be part of a lead enrichment cycle.

Read more

For Tal Fishman, AI was little more than autocomplete a year ago.

That changed in December. Vague prompts that once failed began producing correct results.

Now, AI can turn a plain-English trading idea into a full backtest report that includes data cleaning, code, and analytics, says Fishman, head of fixed income quantitative research at the $248 billion asset manager Lord Abbett.

“The error rate from a vague prompt used to be 70–80%. In December that flipped. In many cases it started working right the first time about 80% of the time,” he told me in an interview.

For Fishman, AI is not infallible, but it makes testing quant ideas dramatically cheaper and faster. Projects that once required weeks of quant time can now be attempted in days or hours.

Counterintuitively, he sees demand for quant work rising, not falling.

“If testing an idea used to take a month, you might say it’s not worth it. But if AI cuts that to a week or a day, suddenly there are a lot more projects you want to do. So far it hasn’t reduced headcount. It’s just increased how much we tackle.”

In our conversation, Fishman discusses:

• Why December’s model releases marked an inflection point for quant research

• How models use internal documentation to reproduce a firm’s research process

• Why cheaper research is increasing demand for quants

• What makes fixed income difficult to systematize and where AI actually helps

• Why some finance professionals underestimate how much AI has improved

This interview has been edited for clarity and length.

Matt: When did you realize how big an impact AI was going to have on your job?

Tal: It was a JPMorgan conference in the city for quants, I think last spring. Prior to that conference, I had started using AI as autocomplete, basically, for coding. The vast majority of the day-to-day work that I do and that my team does is done via code. Its capabilities were starting to slowly get better — it would go from completing a line to completing a block of code, maybe three or four lines at a time.

What I saw at that conference was that Man Group had put on display their own AI model. It was able to go from a very basic research idea — like, “here is a new dataset, and I would like to test whether the momentum effect can be found within this dataset” — and it was a relatively short paragraph that they submitted to the LLM. From there, you push go, and the prompt said something like, “I would like you to produce a backtest report with our usual graphs and tables.” Of course, it was hooked up to a lot of stuff on the backend for them. You push go, and it’s just churning and producing code. They showed a fast-forwarded video of it literally doing everything, and out comes the report. At the time I was like, whoa — if this is real, this is a game changer.

That really changed my thinking from “AI is going to be a type of model we use when we want to do sentiment analysis” to “this is going to fundamentally change how we do our work.” I tried to replicate what they had done, and I think they must have had a really advanced model for that day back then, because I tried and failed to get that working on my end — until December of last year.

Matt: What changed in December?

Read more

Richard Craib runs one of Wall Street’s most unconventional business models: a crowdsourced hedge fund. He also counts JPMorgan as his biggest backer.

Craib, a South African mathematician, launched Numerai in 2015 in San Francisco, far from the epicenter of finance in New York, with the goal of reinventing how hedge funds are built.

Numerai crowdsources stock market predictions from thousands of data scientists worldwide by providing encrypted financial data that obscures the underlying securities. It then aggregates those forecasts into a single trading strategy. Contributors stake the company’s cryptocurrency, Numeraire, on their models, earning rewards for strong performance and losing funds for poor results.

Despite its unconventional structure, Numerai manages real capital and in August secured a commitment of up to $500 million from JPMorgan Asset Management, potentially more than doubling the fund’s size. The investment followed a strong year for the fund, which reported a 25.45% net return in 2024 with a Sharpe ratio of about 2.75.

For much of its history, Numerai framed itself as a hedge fund built by machines but guided by humans.

Craib is now reworking Numerai for autonomous research. Last month, the firm outlined plans to redesign its system to support agents rather than just human data scientists, including a new Model Context Protocol interface that would give AI systems direct programmatic access. Under that framework, agents could create models, submit predictions, run validation tests and monitor performance on their own, effectively executing the full research cycle without manual intervention.

The shift reflects Craib’s view that advances in modern AI tools have changed who, or what, can participate. Human users are expected to move toward designing and supervising AI research assistants rather than building models themselves, while updated staking mechanisms would allow agents to manage financial exposure programmatically.

He expects agents to spread quickly across quantitative finance, potentially reshaping how ideas are generated, tested and traded.

In our chat, we discuss:

• Why Numerai is redesigning its platform for autonomous AI agents, not just human quants

• How large language models became capable of running the full research cycle with the right scaffolding

• Why Craib believes future hedge funds will rely on “AI scientists” exploring vast idea spaces

• How the JPMorgan investment came together and what it signals for institutional adoption

• Why Craib thinks many traditional fund roles, and even star managers, could become obsolete

Here are some of my favorite quotes:

This interview has been edited for length and clarity.

Matt: You started Numerai about 10 years ago, when AI was not as prominent. Now you have JPMorgan investing. How were those first couple of years?

Richard: Actually, I thought when I was starting it, AI was a bubble in 2015. It felt that way. Google had acquired DeepMind for $500 million, which people thought was just really extreme. There was a lot of different kinds of hype at that time, and I guess we were more in the machine learning space, and we weren’t quite on LLMs yet. But that was AlphaGo in 2016, right when Numerai started. But it ended up not being a bubble at all. There was a lot more to come.

Matt: It’s still an unusual model for a hedge fund. Looking at your recent NumerCon announcements, it seems you are setting up the infrastructure for submissions that don’t necessarily come from humans.

Richard: We’ve actually always thought about it that way. When you signed up in 2016 on Numerai, it didn’t say “enter your username,” it said “name your AI.” You were not the one who was doing anything, except setting up the learning algorithm to start learning, and then AI would be the thing submitting. And now that’s become even more true, because even the code that you would write to generate the model, even that code can be written by AI. So, we just see it as another abstraction.

Put it this way, we were never asking data scientists to write machine learning algorithms in assembly code. They were using the most extreme abstractions, so they would use scikit-learn in Python, or TensorFlow, and now there’s another layer of abstraction, which is Claude can do TensorFlow for you, or PyTorch for you.

It’s natural for us since the beginning of ChatGPT since it’s always known about Numerai. It knew how to make a basic model, even on the first version, but then it got better and better. So, users have always been using the chat interface, but we never fully enabled native agent support until NumerCon.

Matt: What made you decide to focus more on this approach? When did it click?

Richard: In November, there was a tipping point that everyone in Silicon Valley felt. Models like Claude and ChatGPT Codex became capable of doing almost anything if you provided the right scaffolding.

That was the moment where it was like, okay, well, now we should really just lean into this, because you get the feeling that everyone will be here soon.

In the beginning of Numerai, there was a popular statistical programming language called R, and that was actually very popular, maybe half and half users used that. But then it moved to Python, PyTorch, almost completely, and I think it’s the same thing with this. You’re just gonna see very quickly people feeling they’re doing it wrong if they’re not using agents.

Matt: So, you see this more as the scaffolding and architecture. From the B2C side, it’s about which model does what, but for enterprises, they’re more concerned with the framework—the tracks it runs on.

Richard: This is the key thing. It’s not super well understood, but if we were to hire a PhD that was super smart, he would still make the basic errors in the first few weeks, because he wouldn’t know how to do a proper cross-validation backtest on financial time series data. And that’s the same with Claude. No matter how smart it gets, it doesn’t have Numerai Skills, which is the skills.md file we made. So, when you ask it, do a whole bunch of research on Numerai, sort of within the first hour, it’ll make three bad mistakes. It’s not really its fault, it’s not because it’s dumb, it just doesn’t know quite how we do things. And so, once it’s got access to the skills.md, it’ll be like, oh, well, if I need to do that, I’m just—I have to use the skills way of doing it. And so that’s how the scaffolding gets defined really nicely.

Matt: So, it’s tailored directions. And it sounds like you’d rather have good scaffolding with a lesser model than a good model without it.

Richard: Yeah, exactly. And the 'Skills' feature was released by Anthropic basically only two or three months ago. If you don’t have that, it’s almost like you haven’t had onboarding. It’s like Citadel University where you spend a month before they let you do anything. You have a whole bunch of training, and so you’re building up skills about how the organization works, how they do things, before they let you touch any production code. So that’s a training camp for AIs.

Matt: Have you heard of Man Group’s AlphaGPT?

Richard: Yeah, I don’t know what it really is, but I’ve seen that they’ve made announcements about it.

Matt: I spoke with Man Group’s Ziang Fang about it. He described it as an end-to-end idea test machine. Whatever passes through certain thresholds, the humans look at it. They’ve said the AI-generated ideas are passing their human benchmarks on the tests—it just lets them test more ideas and do more than they could before. Is that a similar concept to Numerai?

Richard: It’s one thing to say we have good infrastructure to test ideas. But it’s really, can you get to the point where you can span the full space of possible ideas? That’s the trouble with a quant signal. If you think about how many ways there are to order 6,000 stocks—because that’s what Numerai users are doing, ranking best to worst—there are 6,000 factorial. That’s 2 times 10 to the power of 200. There are more permutations than atoms in the universe. So, that’s why we still need the crowdsourcing, because we don’t know what to ask, or what model to build in the first place.

I think if you were 16 years old and you said, “Hey Claude, make me a quant fund from scratch,” and that’s the best prompt you could come up with, it would make you a very generic, mode of the distribution, quant fund infra. Whereas if you say, “I read this paper in 2023, which had these really strange ideas, and here’s the paper, and can you turn this into a Numerai idea in this way,” then that is a much more directed search into what’s possible. So that’s how we see users using this. They still have a role to play in, ‘This is the direction I want to go in,’ and I don’t want something average, because you don’t get paid for an average model on Numerai.

Matt: You’re looking for outliers.

Richard: We literally pay you for orthogonal alpha. So, if you make something crowded, that is the first thing Claude would come up with, and you’ll earn nothing.

Matt: HRT has said at some academic conferences that they’re training foundational models on financial time series—basically the language of financial data. What do you think about that approach?

Richard: We haven’t done that. I don’t think it’s that important. If you just train on the price time series, it’s crazy to say all you need is the price to predict the future. That’s a very 90s thing to say. You are getting paid by the market for adding strange new information to it, not the most commonly known information. So, we have 2,000 features, and not that many of them are based on price. You could maybe cobble something together, but that’s not the be-all, end-all.

The way I see it is more like these models are themselves AI scientists, and they weren’t a year ago. So why not just now run the scientific method more and more?

We have built language models. We built something called Numerai Predictive LLM, and we made it read news and then come up with a prediction from the news. That was an 8 billion parameter model. It actually doesn’t matter if you make it higher in terms of parameters. But that was a natural use case, because the current language models will not be able to predict from news what will happen, because they’re not trained to.

One example we give is, if you have a company like NVIDIA, and they make a press release that says, “we’re being investigated by the Department of Justice for monopolistic practices,” and the second paragraph is, “our revenue grew 150% year over year.” You ask ChatGPT, is this good or bad for the company? And ChatGPT will say, well, it’s neutral, because there’s some good things in it, and there’s bad. And actually, it’s extremely positive for the stock if you’ve fine-tuned the model. So, our model gets that right. It says this is amazing news, and the other models don’t. That’s one place where we are internally building features with language models, but it’s not on the time series of price.

Matt: How did the JPMorgan investment come about?

Richard: The thing about hedge funds is there’s quite a lot of short-term thinking, people want the first 3 years to be amazing. But we were like, let’s not even raise any LP money and raise venture capital, and then build something that no one can compete with in the long term. That’s why it’s been more of a tech company.

JPMorgan, the first meeting with them was something like 2018, 2019. We were probably below $100 million, maybe below $50 million, because they’ve had a lot of success investing in cutting-edge stuff. They’ve invested in early machine learning funds like Voleon and Voloridge, I believe.

In the early days, it was more like us saying, what do you want us to do to be fully institutional and ready for you? And they told us all the things they like. They like the 3-year track record, not a zero-year track record. They, in some ways, helped us make the product something that was institutional grade.

They’re not the first institutional investor—we have 3 endowments and many others besides. But they are the biggest one in terms of capacity, they want to invest $500 million.

Matt: Has that opened other doors?

Richard: Yeah, a lot. We now have a 6-year track record. We look very good compared to peers, and we’re getting better and better. Whereas other peers, maybe they got too big, and now they’re struggling to put up good numbers. But we’re snowballing, where our data is growing, our users are getting smarter, and everything’s kind of getting better, and risk management is getting better. So, yes, after the JP Morgan announcement, many of the big players have been reaching out to us. And probably in the next month or two, there might be other announcements, and we should be over at 1 billion quite soon. We’re almost $600 million, but $600 million is maybe 2 checks away from a billion.

Matt: You also have a different level of auditability and transparency compared to a typical hedge fund. There’s a lot more detail about how models are submitted and staked.

Richard: Yeah, and it’s interesting, because we’ve even had an endowment investor make a user account and submit a model. And he got to really see, okay, this makes sense. I didn’t understand this bit. And also, he got to see that he didn’t win. There were people who were a lot better than him. He sort of saw, okay, this talent is very good here. Anyone can watch our performance. Another thing you can watch that I like to watch is how well the metamodel is doing—the combination of all models—how well that’s doing against the benchmark models, which is just the free benchmarks we give away. But those models, we’ve tried our best to make them very good, so they’re our best internal model. And we said, here’s the baseline model, improve on it. Well, almost month after month, the edge widens, and it’s never looked as wide right now, where the crowd, the stake-weighted metamodel, is crushing the best we can do internally. Because the reality is, we’re good at data science. We have good data scientists. We’ve hired some of the top Numerai users over the years, but we don’t know how to beat everyone. And we don’t think we’ll ever beat everyone even with infinite AI scientist assistance.

Matt: Is that just the wisdom of the crowd?

Richard: No, I don’t even like that term, because the wisdom of the crowd is almost saying that the individuals are dumb, but the crowd is smart. I actually think it’s the opposite. It’s much more like an open source project, where about 1% of the users who’ve ever signed up to Numerai are the core contributors. And then the next 5% is also very important. Numerai, by asking people to stake their models, we are making it hard, on purpose, to do well. If you do badly, you will get your stake destroyed. So, Numerai is more like an API to find the best thousand data scientists in the world, versus let’s all make dumb guesses, and it’ll average out to something good.

Matt: So it’s more like winnowing it down to the best?

Richard: Yeah.

Matt: You’ve been building this a while. What’s been the most surprising thing?

Richard: The one thing I do think is true, and it makes total sense, is that the venture capital industry in this country is just amazing. We raised from the best VCs. Very quickly, they sort of saw a future where it’s like, okay, well, what if the way Millennium works is kind of gonna seem outdated in 2030? Where you hire all these people, and then pay them a lot, and then they read the newspaper and code. It’s weird. But the Numerai way was this new thing, and so it’s always been very easy for us to raise venture rounds. But I would say that the asset allocators, they’re more backward-looking. They say, well, Millennium has a 30-year track record. And we don’t trust AI yet. So that to me was quite tiring, in a way, to basically try to just educate, because when you heard about Numerai, there would often be three things you have to kind of know about. Blockchain, which no one knew about. Then machine learning. And then quantitative finance. So you had to have all three to like Numerai. You had to have a lot of knowledge of all three. And most people were kind of 1 out of 3. Now, I would say people are getting to 3 out of 3, because those are the technologies du jour.

Matt: What about the hype cycle? You’ve seen Numerai get labeled different things over the years.

Richard: Yeah, we’ve been a little bit bubble-averse. There was a similar time where people were talking about us as a blockchain company and hyping up our cryptocurrency. And I was just trying to put some cold water on that, because I just don’t want people to be disillusioned. That’s not really a hedge fund style. It’s supposed to be risk-adjusted, long-run. It’s not gambling.

Matt: Are you seeing more submissions since NumerCon?

Richard: Yeah, it has. NumerCon was less than a month ago, and there’s 150 to 200 MCP connections, and there’s only 500 staked users. They make many models per user. That’s surprising.

Matt: Are you going to get to the point where an agent is working for you and you’re just on the beach?

Richard: That’s the dream, but everybody has agents, too, including Numerai’s peer competitors. I really think that there are pods at Millennium that are paid $100 million a year with code that could be replicated in 40 hours by Claude. And so I don’t know what they do. I think that’s part of Numerai’s mission. We want fewer human beings in the hedge fund management industry. And I think we’ll get there, and Claude is helping.

Matt: What about the broader disruption to white-collar work? A lot of it turns out to be white-collar manual labor.

Richard: I think it’ll be looked back on in almost disgust by grandchildren. With trading, you don’t understand the human mind or intelligence if you think you can go out for breakfast at the St. Regis in New York, have coffee, and then on your way walking to work you’re like, “I should buy NVIDIA.” And then you go and buy it. It’s crazy that you have no information, except kind of the sort of amorphous blob of human thought.

Matt: But that’s not just the retail trader. You’re talking about hedge fund managers doing the same thing.

Richard: I’m worried more that it’s actually hedge fund managers who would be that person having coffee at the St. Regis, and they’d buy $100 million of NVIDIA based on their vibes. And you’re like, you know there’s 2,000 dimensions of data that Numerai has?

Matt: The majority of money managers underperform the benchmark.

Richard: At what point do you realize you had a lucky call, and it had nothing to do with you in some way? It was just an apparition in your mind? We’re very vulnerable to that type of thing.

Matt: Everyone on Wall Street wants to be the smart guy.

Richard: I’m not the smart guy, but I made a website to be friends with all the smart people.

An earlier version of this interview misspelled NumerCon.

Kirk McKeown spent about 15 years running what he calls the “factory”—some of the largest fundamental channel-check and data driven operations on the Street – first, at Glenview, and later, at SAC Capital and Point72. At its peak, his team was conducting several thousand calls each year. Kirk’s role evolved and ultimately, he ran all proprietary research at Point72 across calls and data. After years of managing this massive human-capital engine, he realized the “moat” in institutional finance was shifting from access to data towards the architecture used to structure it.

In 2021, he co-founded Carbon Arc, a platform built to structure data to be sold by consumption, as opposed to locking it up as a long-term asset. Carbon Arc unlocks data trapped on balance sheets and is fresh off a period of rapid institutional adoption. Now, Carbon Arc is betting that the future of alpha resides in a “refinery” capable of structuring 100 trillion transactions for the coming wave of 30 billion AI agents, solving problems not only on the Street, but for all types of businesses around the world.

I spoke with Kirk about his journey from the “manual” Wall Street factory to Carbon Arc’s “agentic” refinery. Here is what readers will learn:

• How the “several thousand calls per year” grind forged a mathematical framework for global market structures.

• Why Carbon Arc treats data as a derivative with time decay (Black-Scholes for data).

• The transition from “drilling” (data collection) to “refining” (knowledge graphs).

• Why the next 12 months belong to “automated agent onboarding” over retail chat.

This interview has been edited for length and clarity.

Matt: You had a long history in the hedge fund industry. What made you decide to jump ship and start something of your own?

Kirk: In 2012, I went to SAC to build what is now called Canvas. I had run a similar business at Glenview Capital, and built a large fundamental research business collecting information in supply chains. From 2006 to 2014, I did thousands of calls a year, myself. That kind of volume created for me strong principles around scaling research problems.

Research frameworks are patterns. The same story happens over and over. What’s happening in the U.S. government right now has happened three or four times. It’s just different names and different clothes. The world follows rules based on market structure, business models, management teams, and personality types.

I started to learn that hospitals and hotels are the same business because they both get paid on length of stay. I learned that TSMC and US Steel are the same business because while they have different end markets, they make money the same way. In the Global 2000, there aren’t 2,000 companies. I believe that there are four market structures and nine business models.

You start to find these scale points. For example, Tractor Supply is like a home center for rural areas. 25% of their business is animal feed and 25% is Texas. If you get a handle on animal feed in Texas, you get a handle on that business and can make a better risk-adjusted bet. I developed mental hacks and mathematical decision frameworks. It wasn’t because I was smart. It was because the “n” was so big.

By the time I left Point72, I was running proprietary research, managing the people responsible for generating actionable insights for the Firm’s investment teams to use as inputs in their process.

While working in research at these great firms, I started looking at the friction in the data market. It traded like 1930s equities: big block trades for bags of cash for market insiders with massive balance sheets. I looked at the legal and compliance frictions—it takes forever to get a data set approved. There were technical frictions. In 2016, Snowflake had been around for two years and Databricks had just come onto the scene. The infrastructure to manage this data at scale didn’t exist before 2010.

The pricing and commercial construct was built for 500 qualified buyers, not five million, so the clearing price was very high. If you could bring supply and demand closer together, smash down the cost of the insight, and sell the insight rather than the asset, you could achieve density and velocity of consumption.

Matt: How has the business evolved since you started it in 2021?

Kirk: We started building in 2021. Fast forward five years, and we have a two-sided consumption-based platform. Data asset owners bring their data, and we structure and graph it for the AI economy. We created composable infrastructure that allows people and agents to plug into the front of the stack. You can hit a modularized data structure to request an entity (like Lululemon), a framework (like revenue growth), or an asset (like credit card data). You compose that element and buy it for $5. We built an ontology that manages the modular analytical framework and a payment processor to meter it.

We started the stack in 2021, but when ChatGPT came along, we recognized we needed to be in graph. We tore down what we had built and started fresh. We rebuilt the platform as a knowledge graph. We have 100 trillion transactions structured in graphs. We modularize the entity structure around companies, brands, people, and locations.

For example, if someone wants to buy the average salary in 40,000 zip codes monthly to understand wallet structure, they can buy that aggregated from us. They don’t have to buy the whole data set for several hundred thousand dollars. I’m making a market for them and running the business like Goldman in the 90s.

Matt: What is happening to the price of data now that the cost of intelligence is decreasing?

Kirk: Companies like FactSet or Bloomberg have valuable data, but they don’t monetize it the right way. The structural commercial relationship between how agents interface with data and how they value it is changing. If you’re selling cases like Westlaw and an AI model ingests it once, they own it. Rewriting that licensing construct is hard because IP rules and laws were not built for agents.

Analytical platforms and database companies get hit because they get paid on compute. Compute is going to be socialized and optimized. It’s the wrong pole in the tent to get paid on. In the oil business, you don’t want to be a driller. You want to own the field or the refiner. Drilling is a bad business. Refining is a fixed-cost, high-volume framework. We are building a refiner.

Matt: How do your former colleagues on Wall Street react to these concepts like knowledge graphs and ontologies?

Kirk: I’ve been evangelizing this for a long time. This is just Wall Street from 1984 to 2025 on a truncated time horizon.

In 1973, the Black-Scholes paper was published. In 1983, Goldman launched the first quant desk, the beginning of the quant age on Wall Street. Between 1984 and 1990, early quant shops competed on models and saw 80% per annum alpha returns. Alpha degraded through the ‘90s as models competed. After the 1998 LTCM crash, ETFs formed because quants needed bigger liquidity profiles. Following the 2007 quant crash, factors came along, rates went to zero, and the factor market formed. Over that time, commissions went from $2 a share in 1983 to less than a penny today, while volumes went up 10,000x. In 1984, 50% of New York Stock Exchange trades were blocks. Today, it is 7%. Our stack is built to remove frictions to allow models to engage.

In 1985, when models started to proliferate, the traditional guys poo-pooed them. In 1995, when electronic trading came along, floor traders said it would never work because people liked talking to people. It’s the same thing as Blockbuster. There are resistors, but everyone uses data.

To us, OpenAI and Anthropic are just hedge funds. They are writing models to create lift in decisioning and competing on that lift. They are buying and selling scientists the way Millennium and Citadel do. I’d argue they are on the wrong capital structure—they should be raising GP/LP stakes rather than VC money. They don’t have a moat other than capital. Wall Street ends up winning the AI wars over the medium term because of regulation and their historical relationship with modeling the world.

When I worked at Point72, I had to find simple analogies to manage a big group. Data is a content business. You can’t own data end-to-end as an individual. You need engineers, scientists, analysts, and salespeople. Content must be relevant, differentiated, and accessible. Accessibility is asymptotic and relevance is table stakes. Differentiation is the only thing that separates us, and in content, that means more data and better questions.

OpenAI and Anthropic have trained on a relatively small amount of data, mostly scraped from the web. To manage the world’s inventories and inform global decisions, you need access to transaction data that shows how people spend their time and money, and their balance sheets. That is what Carbon Arc has built. We have 75 assets, three petabytes of data, and daily granularity for $150,000 a month in compute. We smashed the cost down. Now we are scaling both the supply and distribution sides.

Matt: Can large hedge funds or banks build this themselves, or do they face structural issues?

Kirk: They can build it, but they have a competitive problem. They monetize data through the market, so they won’t share their alpha back with data providers. We built a data transaction processor that creates liquidity for data providers and opens up their distribution. Data providers are coming to us because they can distribute broadly rather than doing one big exclusive check with a firm like Two Sigma, Citadel or D.E. Shaw.

Hedge funds want exclusive data and don’t want it proliferated. Data providers just want to get paid. Because data has historically been expensive and hard to work with, only global businesses and large funds could buy a million-dollar data asset. That market structure is what we are changing.

We launched platform 2.0 in mid-2025. We started last year with 35 customers and ended with 75, quadrupling revenues. Half of our customers are Wall Street buy-side and sell-side. We work with five of the top eight consulting firms, and we have good coverage in media and Hollywood. Companies like Paychex are both suppliers and customers. We are launching automated onboarding for agents on February 17th. We didn’t build this platform for eight billion people. We built it for 30 billion agents.

Matt: How do you see the market for small and mid-sized businesses (SMBs) and retail users?

Kirk: We are launching retail in March 2026. We will launch our MCP server for people with Robinhood or Kalshi accounts. We’re going on Reddit to offer a hedge fund data stack for $20 a month. For SMBs, a VP of Finance at a small healthcare business can pay $200 a month to do competitive analysis by plugging their Claude into our stack via the MCP to query credit card, paycheck, and healthcare claims data.

A consumption-based model needs volume. We give away publicly sourced data, like SEC data, for free. This is a cost game. Models are democratizing analytics. If you are building on top of public data and overpricing it, you will lose. We think about the business in terms of cost per megabyte and price accordingly.

Markets are forming for things that seem bizarre, like Kalshi’s contracts, but the real differentiation is composable contracts on anything, anytime, anywhere. We are in the third inning of a doubleheader. This technology cuts friction and makes things economically viable that weren’t before.

I am an AI bull, but I am concerned about the next 10 to 15 years. The dislocation in the labor market will take time to absorb. There are massive regulatory and ethical issues. Civilization-changing situations—like electricity in the 1880s or the rise of quants in the ‘80s—always involve these cycles.

Matt: Where does the “moat” for your business exist in the long term?

Kirk: The moat ends up being regulatory. Our General Counsel came from Schulte Roth & Zabel, the largest data compliance shop. She is standardizing compliance as a product. As agents proliferate, we must ensure legal and regulatory standards are met. Right now, it’s the Wild West, with major publishers suing Silicon Valley shops for scraping.

We are leading with scalable compliance frameworks. It’s like Stripe. You need regulatory infrastructure to scale. We model the business after Goldman Sachs. Someone once said to me that Goldman is a regulatory wrapper that allows you to do cool stuff in 100 countries and apply capital against it. They are an enablement platform that marries regulatory access and capital.

In the future, someone will emerge as the “Moody’s of data,” scoring inputs. Centralization will happen around core scale points, just as it did with Coinbase in blockchain. Maintaining a competitive advantage when things move this fast is hard. It’s an infrastructure build. Some people stand up application layer businesses on top of OpenAI and hit $10 million in revenue in six months, but that’s a gold rush. It’s not sustainable because there’s no underlying moat once others join. We are building the infrastructure. We think that scales. We think that sustains. We think that’s permanent.

We aren’t going anywhere any time soon.

Norway’s sovereign wealth fund runs the biggest pool of capital in the world.

And its CEO, Nicolai Tangen, might just be the biggest advocate of AI in investing, calling himself a “total maniac” about it.

Stian Kirkeberg is tasked with implementing Tangen’s vision across roughly $2 trillion in assets and about 8,600 companies as NBIM’s Head of AI and ML.

That scale brings a specific set of constraints: broad market coverage, strict ethical rules, and an organization that has to work reliably across thousands of decisions. An individual can quickly boost their own output with vibe-coded solutions, but that does not necessarily translate into a faster organization. When everyone becomes a coder, productivity can rise in pockets while technical debt quietly accumulates.

In this interview, Kirkeberg walks through how NBIM is navigating this transition. He explains their partnership with Anthropic, the move from a bottom-up ambassador model to a more centralized strategy, and how small autonomous teams are replacing traditional Scrum structures. He also gets specific about how they reserve GPU capacity from hyperscalers and how LLMs are being used to screen thousands of companies for ESG compliance.

By reading this conversation, you will understand the constraints that show up when AI-driven development scales, and why the biggest hurdle to ROI is not the model’s performance, but the organization’s ability to absorb what it produces.

This interview has been edited for clarity and length.

Matt: How did you get connected with Anthropic?

Stian: Last autumn, Nicolai invited Dario to his podcast. From there, the ball started rolling. While we were evaluating which tool to buy, Anthropic came out on top. At that time, it was OpenAI and Anthropic, and the others weren’t that great.

Matt: It took a decade for the move to the cloud to happen. This technology is still relatively new.

Stian: We were really fortunate to get this collaboration with Anthropic. We started with basic training and prompting for everyone. Then we set up an AI Ambassador Network which grew from 20 to over 70 people. My AI team had meetings with Anthropic twice a week. Ambassadors were tasked with finding a use case in their area, solving it with the AI team, and then showcasing it to the rest of the organization.

We built a lot of momentum with success stories. This was umbrellaed under “Tech Year 2025.” We created mandatory training for everyone in NBIM—seven different modules covering prompting, critical thinking, and responsible AI. We rolled out Claude, Cursor, and Copilot for everyone who wanted it. We had internal conferences in each office where people celebrated good stories and brought in speakers. We even had a Neo1 robot from a company called 1X.

After that bottom-up approach, we needed to identify the most valuable use cases for NBIM as a whole. Consultants interviewed the chiefs and ran workshops, identifying another 171 projects. Phase 3, which we focus on now, is about people delivering value on everything they’ve learned and the tools they’ve been given. We are pushing the cultural change this year to show the value of those investments.

Matt: In practice, where have LLMs proven most useful?

Read more

When generative AI began gaining traction on Wall Street, many firms responded cautiously, often firewalling off the technology from its employees. At Manulife Investment Management, the reaction was different. After years of investing in cloud, data, and machine learning infrastructure, the firm moved early to establish an AI framework across the organization, building governance, risk controls, and a process for prioritizing use cases.

I recently interviewed Robi Krempus, who leads AI for global wealth and asset management, which has $1.3 trillion in assets under management and administration. Earlier in his career, Krempus was a control systems engineer in the energy sector, working on nuclear power thermodynamics and other high-stakes modeling problems. That background now shapes his role overseeing Manulife’s AI platform. Rather than committing to a single vendor, his team has built a model-agnostic framework that allows the firm to move between providers such as OpenAI, Anthropic, and Google as the technology evolves.

In our chat below, Krempus explains why Manulife moved quickly, how his team co-designs tools with portfolio managers, and why the firm shifted from project-based experimentation to a platform strategy. He also discusses how Manulife evaluates large versus small language models, how it manages tech debt as models change, and where AI is already proving useful, particularly in extracting qualitative signals from standard financial disclosures.

This interview has been edited for length and clarity.

Matt: Manulife moved quickly when generative AI first emerged. At the time, many big firms were banning or firewalling it. What drove that decision?

Robi: We truly saw the opportunity. We had already established a strong data science and machine learning community at Manulife. When you build traditional machine learning models, it is often about forecasting or predicting a variable, which still requires a massive infrastructure. When generative AI came into the mix, we quickly understood that this is much broader and will impact everything—decision-making and how you think about intelligence. Organizationally, we saw the opportunity, and with the CTO, CIO, and Chief AI officers, we were certain this technology was not going to go away. Unlike emerging technologies like blockchain that take time to embed, it was quite apparent that this would be transformational.

Matt: How do you architect this technology? How do you organize it to get started?

Robi: In asset management, there were three ingredients where we believed this would really make a difference. One is strong leadership support. We had huge support from Colin Purdie, the Global Chief Investment Officer for Public Markets, and his leadership team. Secondly, we co-designed solutions with the investment professionals. We have CFAs on my team, but we are not managing the money; the investment professionals are. That co-design allowed us to tackle specific pain points together.

Thirdly, our mindset shifted from being project-based to a platform mindset. We wanted to establish a platform so that whenever we have an additional use case, we can give AI to the end user through that platform. We have seen adoption over 70%, and we hold weekly office hours where investment professionals can stay on top of new features.

Matt: Some firms use various models as an engine and build an application layer on top. Can you walk me through your thinking on building those applications?

Robi: From a Manulife perspective, we have a robust model risk management process in place. Before anything goes into production, it is vetted against hallucinations and quality. In working with investment professionals, quality matters a huge deal. If the LLMs do not produce an output that hits the investment context, it will not work. We architected our AI with feedback loops and tested various systems to increase output quality and reduce hallucinations. It is not a straight-through process to a reasoning model; spending time on the AI architecture to increase quality was really impactful.

Matt: Are you agnostic to the model? Can you swap different models in and out of your infrastructure?

Robi: Yes. That goes back to the ten years of investment we put into infrastructure and cloud. What is amazing now is the availability of all these models. Even when OpenAI released 3.5, we had access to it quite fast. The idea was to create a data framework that allowed us to productionalize models in a responsible way. We have a broad lineup available, whether it is OpenAI, Anthropic, or Google. It is fast and responsible.

Matt: What are the most common use cases right now, and how have they evolved?

Robi: We started with discrete use cases. One that seems obvious is earnings call transcription. We co-designed solutions with investment professionals to build standard prompts for things like red flags, concerns, or bull-and-bear situations. This was managed in a prompt library and helped support investment conviction.

What we did next was aggregate that data. If you take earnings calls and add outside reports or notes, it allows you to search across the board. You can search across your portfolio or a sector for specific topics. That has been very helpful for deeper intelligence across coverage. We also use it for sustainability, which is an efficiency play to quickly get information out of very long documents.

Matt: How are you thinking about small models versus larger ones?

Robi: We have teams that constantly task new models out. We have looked into small language models for operations or distribution areas and have seen a fit there. We are not deploying small language models into asset management right now because we are very pleased with what we can do with large language models. Organizationally, we work strategically on small language models regarding cost and scale, but it hasn’t impacted asset management yet.

Matt: How do you decide between building something internally versus using a third-party source?

Robi: We look at it as “buy, build, or reuse.” Because Manulife is a large organization, we first see if we can reuse something already built. We have a stream that constantly evaluates vendors to see if a solution makes sense. The last thing we want to do is manage internal tech debt. In some cases, we bring in vendors; in others, we build. The platform mindset matters here because it reduces tech debt while allowing us to fine-tune and differentiate ourselves in the marketplace.

Matt: What do you think is currently overhyped or underhyped?

Robi: I am particularly curious about autonomous coders in the bigger tech space. In the past, if you built a machine learning model and a new algorithm came out, you had big expectations for accuracy, but you still had to do so much feature engineering to improve it. Now, innovation and design matter because you have so many options in how you architect data and AI.

Regarding what is overhyped, the real impact is AI’s ability to deeply analyze structured and unstructured data in an automated way. In asset management, with the enormous amount of qualitative and quantitative data, that is where it gets interesting. When these things merge toward AGI, AI will be able to figure out insights and analysis from any sort of data using natural language. You won’t need to know Python to dig that information out.

Matt: How do you see AI impacting alternative data sets and how people use them?

Robi: There is a progression and a cultural change involved. Our mindset was not to wait and establish a perfect, integrated, scalable data infrastructure before building AI. Instead, we put AI into the hands of end users to learn and adjust based on feedback. While we haven’t fully tackled alternative data sets in our AI journey yet, there is an opportunity within standard data sets. For example, in sustainability reports or the footnotes of documents, there can be instrumental nuggets that take a lot of time to find manually. We ask ourselves how we can get critical information out of the data sets we already have readily available.

Matt: What has been surprising to you in building this out?

Robi: The agility really matters. Two years ago, ChatGPT 3.5 came out, and now the world is talking about the AI workforce and humanoids. What was initially a surprise is how fast you have to rethink things. We might build a solution in January, and then a new model like Claude comes out which is an excellent execution engine. The surprise is the constant reimagining and adapting. If you asked me a year ago, I would have been surprised by how far we have come.

Early users span hedge funds, insurers, and sovereign wealth funds. Bridgewater has used Claude to help researchers query internal documents and data. AIG has applied it to underwriting and risk analysis. Norway’s sovereign wealth fund, NBIM, uses it to work through policy and investment material at scale.

I was happy to speak with Jonathan Pelosi, who leads Anthropic’s financial services effort, about how firms are actually using the product. Here is what readers will learn from the conversation.

• How Anthropic is tailoring large language models for regulated financial workflows

• What Claude for Financial Services is designed to do in day-to-day finance tasks

• How skills and Model Context Protocol connect models to firm-specific workflows and data

• Why…

Read more