Nice - my AI startup school talk is now up! Chapters: 0:00 Imo fair to say that software is changing quite fundamentally again. LLMs are a new kind of computer, and you program them *in English*. Hence I think they are well deserving of a major version upgrade in terms of software. 6:06 LLMs have properties of utilities, of fabs, and of operating systems => New LLM OS, fabbed by labs, and distributed like utilities (for now). Many historical analogies apply - imo we are computing circa ~1960s. 14:39 LLM psychology: LLMs = "people spirits", stochastic simulations of people, where the simulator is an autoregressive Transformer. Since they are trained on human data, they have a kind of emergent psychology, and are simultaneously superhuman in some ways, but also fallible in many others. Given this, how do we productively work with them hand in hand? Switching gears to opportunities... 18:16 LLMs are "people spirits" => can build partially autonomous products. 29:05 LLMs are programmed in English => make software highly accessible! (yes, vibe coding) 33:36 LLMs are new primary consumer/manipulator of digital information (adding to GUIs/humans and APIs/programs) => Build for agents! Thank you again for the invite @ycombinator and congrats again on an awesome events! I'll post some links/references in the reply.

Founder's Journey: The Problem with Solutions Call me crazy, but after spending the last 17 years as a Venture Capitalist, I've finally returned to my roots and am sitting once again in the Founder’s seat. At the ripe old age of 54, I've traded my seat as an Investor for the rollercoaster ride in the Founder's seat. Consider this post the first in a series of field notes from the trenches: Equal parts confession, revelation, and therapy session. Today's revelation? The asymmetry of certainty between problems and solutions. Let’s start with a very basic framework that I used every day in my role as an Investor. It didn’t take hearing thousands of pitches to realize that at the core of each Startup is the pairing of a problem statement and a solution statement. The problem statement is the Founder’s way of explaining a problem they’ve discovered and an articulation of why it’s a gigantic and profoundly painful problem. The solution statement is the Founder’s way of articulating a better way of solving the problem they described in the problem statement paired with a narrative explaining why their users are likely to agree. But I’m reminded of something both obvious and profound now that I’m back in the Founder’s seat: Even though I can present my problem and solution statements with equal story telling elegance, the certainty behind the problem and solution statements aren’t close to the same! When you spot a problem worth solving, it's like discovering a new continent. You can map it precisely, document its features exhaustively, and articulate exactly why it matters. The problem statement becomes your north star: Clear, compelling, and unwavering. But solutions? They're mirages. What looks solid and complete from a distance reveals itself to be merely an outline when you get closer. That brilliant solution you're ready to shout from the rooftops? It's mostly wishful thinking wrapped in confident packaging. I find myself in this exact position as we speak. My problem statement feels bulletproof. It’s backed by research, validated through countless conversations and supported by data. My solution seemed equally solid...until I started peeling the onion with my co-Founder and early ride-or-die team. Every day reveals new questions and problems I haven't considered. The elegant solution I’ve been pitching feels like a child's drawing compared to the architectural blueprints it will need to become. The temptation is to keep the illusion alive and to maintain the facade of certainty when speaking to potential team members and Investors. After all, who wants to follow a Leader into battle who admits they're not entirely sure where they're going? But the truth is that we need to find talent that embraces this uncertainty. They’ll need to understand that the initial solution is merely a hypothesis. The real solution will emerge through iteration, experimentation, and the collective wisdom of a team looking at the problem through different lenses. And even then, it’s only the starting line for the ultimate reality check: Users. They'll take our precious solution and use it in ways we won’t have imagined and highlight flaws we’re blind to. That’s when the real work begins. This journey from problem certainty to solution discovery isn't a bug in the founding process: It's what it’s all about. It's where the magic happens, where true innovation emerges, and where I’m hoping to earn some new battle scars as a Founder. So my advice? Fall in love with your problem statement, not your solution statement. Be humble about what you know and honest about what you don't. And remember that the gap between your initial solution concept and what eventually works is where the real journey takes place. Onwards and Upwards. April 28, 2025

Watch our recent webinar on how to optimize your embedding model using Topological Data Analysis (TDA)

.@bluelightai Cobalt is now available worldwide on PyPi! Data scientists around the world now have easy access to the powerful model illumination and improvement capabilities in BluelightAI Cobalt. Try it today by typing “pip install cobalt-ai” into a Python shell or notebook.

Jimmy Carter sold his peanut farm, but the owner of a major defense contractor and tightly regulated car company is heading up a review of regulators.

Tomorrow night through Monday, I will be giving a series of talks throughout Pennsylvania. If you’d like to attend one of my talks, there’s no attendance fee. You just need to have signed our petition supporting free speech & right to bear arms & have voted in this election.

If I bring up Trump in LA, people react like they got shot with a dart containing rabies and crystal meth 😂

The incitement to hatred and violence against President Trump by the media and leading Democrats needs to stop.

Founder Mode, done right (thread):

Wow