AI Isn't a Personal Tool — It's an Organizational Change Leaders Must Model Themselves
AI is not a personal tool; it is an organizational change that leaders have to demonstrate personally. At Walleye, everyone uses AI every day and every meeting gets recorded, building a first-mover advantage out of operating leverage and a data lake.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
The turning point was one person's demo, not a strategy announcement
In March 2023, a TMT analyst built a workflow on top of GPT-3, which had only just appeared, making himself far more efficient as an analyst — possibly efficient enough to replace himself. Will was fairly skeptical at first, because this was not how things were typically done; but after seeing the demo he was convinced that "this is the direction." Walleye launched an internal AI project right after, and was early to trying agents as analytical support inside fundamental stock picking. What the example shows is that the real turning point is not a grand declaration, but the demonstration effect produced when someone who understands the business applies a new tool to their own job.
— WillNot using AI is like refusing to go online in 1995
Will sent a firm-wide email with the subject line "AI at Walleye." It opened: "I wrote this email with ChatGPT, you should too, and you should be proud of it"; and, "not using these tools is like refusing to use the internet in 1995 because it wasn't perfect." He defined AI outright as not cheating, but as a magic potion that makes everyone 20% smarter. The email also laid down a rule: anyone who has to write, research, analyze, build decks, handle data, or think for a living has to use AI every day; managers have to get fluent themselves first and then push their teams, because that is part of their actual job.
— WillUsing AI is only cheating inside academia, not in business
Will has watched plenty of people quietly use ChatGPT to write their emails and then pretend they had merely polished the text themselves; he thinks this is stupid self-deception. He quotes Jim Collins to say that AI is about "building the clock, not telling the time": it frees people from mechanical writing so they can think about the next-order task. He draws an explicit line between academia and commerce: in academia, using AI is cheating; in the business world, results are what matter. In the future, even if you manage no humans at all, you will be managing a set of AI employees, and that is a skill in itself. So the point is not shame — it is operating leverage.
— WillWhat AI outsources is the rhetoric, not the thinking
The host worried that letting an LLM ghostwrite would make thinking shallower. Will says the key is separating "concept" from "language and syntax": a large share of writing time used to go into decoration — avoiding ending a sentence on a preposition, making a sentence sound smarter — and that is exactly what AI can take over. His own method is to type out the points and the logic himself, to be clear about "what concept am I trying to convey," and then have the model expand it in his voice. A task that takes 15 minutes used to take four or five hours. The time saved does not go into knocking off early; it goes into thinking more deeply, because AI will not think for you — it has only outsourced the rhetoric.
— WillA hedge fund has to become one collective, not a group of individuals
Walleye's long-term goal is to become a "collective": pour all of the firm's information — email, Slack, call recordings, reports — into a data lake, then use AI to extract insight from it. The most mature case is an internal product called Current: it structures analyst notes, sell-side reports, PDFs and earnings call transcripts, so that every PM can work through all of that text in real time during earnings season. Will says more than 50 outside firms have already applied to join the beta. What machines can do now is not just summarize, but read, listen to audio, and perform second- and third-order processing — and that is displacing the work of human information synthesis.
— WillWorld-class firms will still need thousands of employees, most of them machines
Will draws an analogy to the upheaval between the American Civil War and World War I: railroads and the transatlantic cable changed the world completely within a few decades, and many people went from having obsolete skills to being pushed out inside their own lifetimes — and this round will be faster. His judgment about the future is that a world-class company will still need thousands of employees, but a very large share of those employees will be machines, and there will be far fewer humans. He stays cautiously optimistic: just as the cowboy was eventually absorbed by the "civilization" that barbed wire and the railroad brought, technological inflection points always meet resistance from the old order — but what leaders need to do is not to defend, it is to lead the embrace of the change.
— WillHuman intuition is a neural network too, just one you can't explain
Will points out that quant investing has been "going full speed" since at least the early 1990s: models hunt for statistical structure in stock prices and information, and much of that structure is already beyond human understanding, much like the non-linear relationships inside an LLM. He still believes human investors will keep an edge in novel situations where "the law of large numbers does not apply," because the machine has not seen every precedent, and humans are better at handling ambiguous situations. He takes it a step further and compares human intuition to a neural network: both are high-dimensional judgments built on an enormous base of experience, impossible to explain precisely yet trustworthy. AI's role is not to replace intuition but to act like a coach, prompting you that "you didn't notice that these two moves are actually connected."
— WillNobody skipped journaling out of laziness — it simply cost too much time
Will uses AI to keep a journal: every day he goes through three sections — family, work, health — says a few key points, and AI fills them out into an entry in his voice, the whole thing in under a minute. It started because in finance he could look back at his daily P&L but could not remember what he had been thinking at the time, so he began keeping a record. He thinks most people never journal simply because it used to cost too much time, and AI has driven that time cost to nearly zero. It is also his "you can't manage what you can't measure" principle in practice: he logs every workout and his body metrics, and watches the long-term trends. His two most important first principles are "the power of incentives" and "intellectual honesty" — the first for understanding what motivates people and organizations, the second for facing real decisions.
— WillIn their own words · checked verbatim
Using chat TBT is not cheating That's a non applicable idea from academia.
Ben0:00
I use Chat ET to write this email, you should be using it too and be proud of it
Will14:19
a lot of those employees are just going to be AI robots and that's a skill of in and of itself
Will18:58
these tools don't negate the necessity to think
Will34:23
history doesn't repeat itself it rhymes like that's absolutely true for humans
Will50:13
I keep a journal you know every single day with AI now
Will57:52
Figures
| Walleye assets under management | close to $10 billion | 0:00 |
| When the full pivot was triggered | March 2023 | 6:42 |
| Time a memo used to take | four or five hours | 26:04 |
| Time to draft with AI now | 15 minutes | 26:04 |
| Outside firms applying to test Current | more than 50 | 37:51 |
| Time for the daily journal | 1 minute, sometimes 30 seconds | 1:02:13 |
Glossary
- Borg
- A collective-intelligence concept from Star Trek; here, the state in which all of a company's information flows into one data system that everyone shares.
- operating leverage
- Using tools or systems to amplify individual output, so the same input produces more results.
- pass-through structure
- A hedge fund tax structure in which investors bear the gains and losses directly, giving the fund flexibility in how it spends.
How to listen
Mid- and senior-level managers currently pushing firm-wide AI adoption, hedge fund and quant practitioners, and knowledge workers worried that using AI might be cheating.