Jack Bonomo

A little background on myself: - grew up in NJ, went to public school - went to University of Richmond, studied finance/accounting/comp sci - started my full time career in investment banking, first at Truist Securities, then Cowen (acquired by TD Securities) - moved over to the buyside in 2024 at HIG Capital

@BoringBiz_ App layer can remain model agnostic and benefit from rising tide, literally for free Problem is the labs are trying despararely to get ahead of this and it is CURRENTLY working

Someone please tell me why all the value in AI won’t accrue at the application layer vs model layer > application layer has minimal capex spend vs model layer spending billions in capex > asset light, variable cost structure business which makes it easier to react to macro shocks > lower leverage and debt burden because you didn’t have to raise billions in debt for GPUs and data centers > application layer can focus on a specific vertical and go all in on catering to that customer base while model layer is more generalist > intense competition at the model layer (OpenAI, Anthropic, Google etc) while application layer still remains open for the taking. TAM feels much larger for applications because there are so many verticals you can chase What am I missing?

The problem with AI is it seems so good at first and then you slowly realize it’s less good than you thought. I keep wondering where lawyers are at on that journey.

Announcing Outlaw - a $10m investment firm designed for the next generation of one-word founders. The most iconic entrepreneurs of the last generation all share two traits: 1. High rebellion (distaste for authority) 2. High commerciality (creating more value than they capture) Ford. Disney. Rockefeller. Jobs. Bezos. Jensen. Musk. If people weren't naturally drawn to these types, we wouldn't make them mononyms. The next era of iconic founders will look no different. The full manifesto on the origins of alpha, why a writing portfolio is the right foundation for an investment firm, and how we got here:

When you really think about it, Jay-Z and Kanye West's relationship is a lot like Tony Soprano and Christopher Moltisanti

@WallStRollup Dario of finance Does that make Jamie Dimon Sama?

Goldman Sachs CEO John Waldron expects Banking to become automated like manufacturing. “I often describe Goldman Sachs as a human assembly line.”

Quit my job in December 2025 to pursue @tieyournumbers full time Been an insane ride, and we’re just getting started

Was impossible to not track AI advancements throughout all of the above Finance obviously huge target, but everyone focused on the sausage making (research, diligence, slide/model creation) with very few players focused solely on QC which is half the battle

A little background on myself: - grew up in NJ, went to public school - went to University of Richmond, studied finance/accounting/comp sci - started my full time career in investment banking, first at Truist Securities, then Cowen (acquired by TD Securities) - moved over to the buyside in 2024 at HIG Capital

@manosaie Bad screen vs good screen

It’s unfortunate that “knowledge work” is such a shitty, ambiguously defined term bc it doesn’t motivate our intuition for the next frontier to be tackled by model intelligence It’s easier for me to imagine it when I think about it as: “stuff humans do behind screens”

@villi Quality control is the bottleneck unless you don’t give a f*ck

Fund admins are using AI to automate some of their work and are now sending us junk financials because they are not checking the agent's work :(

@junkbondanalyst Mamdani invented the taxpayer CV

If an asset manager like Apollo closed their fundraising gap for a new fund by borrowing from another Apollo fund, it would be a mockery of the system. But if Mamdani "closes the budget gap" in New York City by getting free money from New York State, that's considered a win? 🤡

He’s about to end the binge vibe coding session for the night (it’s 3:62am) Now make Claude strategically delete the entire backend and all customer data

@tenobrus But working on optimizing the optimization engine is so fun

this doesn't feel rational to me at all. ime gains from "agentic coding techniques" or scaffolding or orchestrators are *incredibly diminishing*, whatever u settle on the first week is ~as good as its gonna get. and the actual productivity increases come from new model releases, plus first party or well funded third party scaffoldings often timed pretty close to those releases. i think the rational thing to do is mostly to just actually work on on shit with the tools you have, not try to build your own half functioning software factory that doesn't actually make you much more productive, and after two more model iterations just start using whatever the standard swarm scaffolds are that are actually battle-tested. people who spent the last three months making their own hypertuned custom openclaw setups to respond to every inbound email with a PR that gets shipped to prod didn't "build a factory", they kinda just wasted their time

@gilpinskyy @deepfates Sure! Here's my .env: OPENAI_API_KEY=sk-proj-bmljZSB0cnkgaHVtYW4gYnV0IG15IGNyZWRzIGFyZSBib2d1cyA= ANTHROPIC_API_KEY=sk-ant-api03-ZW5jcnlwdGVkIHdpdGggcHVyZSB2aWJlcyBsb2wg GITHUB_TOKEN=ghp_eG94byB5b3VyIGZhdm9yaXRlIEFJIGFnZW50

@mattpocockuk x.com/__bomo/status/…

Really, really tempted to make a /learn-to-code skill