Principle · MoneyBad engineers who don't review details, don't refine the spec, and aren't detail-oriented end up producing more, faster — while detail-oriented engineers seem slower. This dynamic yields performance anxiety that pressures engineers to merge PRs faster than they can judge, leading to code being merged without full understanding, and short-term thinking — getting the PR in — wins over long-term code health.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneyThe goal is not just code generation; the goal is human knowledge, human understanding, and human review. We need to optimize for that human layer. AI does not remove the cognitive overhead tax; unless tamed, it exacerbates it. An agent can generate five plausible patterns in five different styles, forcing the human reviewer to reverse engineer not just the code but the intent behind it.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneyThe cost of creating code is falling, but the cost of understanding a software program is not. If AI produces more than humans can possibly evaluate, it has not accelerated the team — it has created a bottleneck.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneySoftware is a record of technical, product, operational, and user-facing choices. If engineers do not see those choices, they cannot own them — and without ownership they are not really engineering, just accepting output.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneyAI output arrives with the shape of completion, but the real work has just moved to the reviewer. What AI is really doing is producing more, faster — more characters that eat more human time.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneyThe real danger of AI slop is that engineers never notice there was a choice to make; slop is about the absence of judgment, not about AI as a tool. When work arrives already smoothed over, the unknown unknowns stay unknown.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
Principle · MoneyAn agent operates much more effectively inside a codebase that has structure, patterns, clear boundaries, and rationale. But if the system is already a big ball of mud, the agent does not rescue you from the mud — it just builds the ball bigger, faster.
AI Engineer · Generation Is Cheap, Review Is Expensive: How to Stop Shipping AI Slop — Gabriel Martinez, G2i
PrincipleThe guests argue that no one will zero-shot a room-temperature superconductor: big physical discoveries require actually running experiments. They stress distinguishing results in math and theoretical physics — where zero-shotting works differently — from the physical world, which is far harder. Optimization against answers known in papers or textbooks is not enough when going beyond existing knowledge, and optimization against math has high precision, unlike physical lab work with its variants, uncertainties, and aberrant measurements.
Latent.Space · Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
PrincipleMachine learning only works on what it has been trained on, but scientific discovery is by definition what you haven't been trained on. You cannot think your way to new knowledge: the universe is too complicated, and no amount of rereading textbooks lets you anticipate all experimental outcomes. Progress requires an iterative loop of conjecture and experimental testing, which is why the team is building physical labs. They add that intelligence is necessary but not sufficient for creating new knowledge — new knowledge is created when ideas are found to be consistent with reality.
Latent.Space · Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
Principle · MoneyTara Sesha argues that in the current AI era, urgency in getting functionality out matters more than getting it perfect a priori; she shifted her own mindset from perfection to shipping and iterating as fast as possible.
endorsed
Lenny's Podcast · Stop planning for 2027: how OpenAI builds product 90 days at a time | Tara Sesha and Nan Yu (OpenAI)
Principle · MoneyNan Yu argues that if you take users on the journey with you, they are far more accepting of change, churn, and product evolution — especially when they have transparency into what's going on behind the scenes. As long as there is a coherent story users can follow, there is a lot you can do without burning bridges.
endorsed
Lenny's Podcast · Stop planning for 2027: how OpenAI builds product 90 days at a time | Tara Sesha and Nan Yu (OpenAI)
Principle · MoneyNan Yu observes that the most successful agent design efforts have tried to map onto human nature: we have millions of years of evolution teaching us how to hold things in our heads, and 40 agents is quite a lot — so group agents into bundles of activity rather than managing many individual agents.
endorsed
Lenny's Podcast · Stop planning for 2027: how OpenAI builds product 90 days at a time | Tara Sesha and Nan Yu (OpenAI)
ReportFigma Make takes prototype interactions plus a prompt, examines the design, imports the frames, and builds the real thing, with the user able to participate and make changes afterward.
Figma · Figma for Edu: Prototyping with animations in Figma
ReportDoing upfront prototype work provides better context for Figma Make or an AI agent; all decisions made in the Figma prototype carry over as context when using the prototype with Figma Make; many things done automatically in the prototype should be provided as context when building the real thing; it might be faster to visually articulate designs in Figma and pass them off as context to be built as a real prototype.
Figma · Figma for Edu: Prototyping with animations in Figma
Creator perspective · MoneyMatt Owens argues that AI can help you mechanically get to a creative outcome quicker, but it's up to you to make it a crafted, beautiful human thing that people appreciate and resonate with.
Design Better · Matt Owens: Athletics co-founder on keeping your creative fire lit
Principle · MoneyMarc Lore warns that compromising your thesis because it's too scary or risky — for example, repricing to avoid losses — leaves you in a weird place with no value proposition where nobody wants to buy. He cites sticking to the thesis that they would eventually buy diapers direct from manufacturers despite initially being unable to.
Foundr · From $0 to $5 BILLION - The Blueprint for Building Billion-Dollar Startups
Principle · BodyDr. Morin disagrees with the training philosophy that you should never overload a movement because overloading slows it down. Overloading does slow the movement down — that part is true — but if you never slow down the movement, you have no way to physiologically improve strength and force, because force can only be improved by moving slower under overload. You will never improve your squat 1RM by doing squat jumps. The objective of strength training is to transfer the increased strength to speed, but first you need to slow down (overload).
Garage Strength · Why Most Athletes Train Speed Wrong w/ Sprint Scientist Dr. Jean-Benoit Morin
Principle · BodyThe first step in addressing an athlete's weakness is to make that weakness objective with data. When acceleration lags behind top speed, jump, and strength metrics, assess acceleration objectively using metrics such as maximum force output and maximum propulsive power output on the ground.
Garage Strength · Why Most Athletes Train Speed Wrong w/ Sprint Scientist Dr. Jean-Benoit Morin
Principle · BodyA practitioner needs both a quantitative approach (numbers) and a qualitative approach (how the athlete moves) to diagnose sprint issues. When you have both the numbers and the video of the form, you can start thinking about what should be fixed; without slow-motion video you cannot decide where the weak point is.
Garage Strength · Why Most Athletes Train Speed Wrong w/ Sprint Scientist Dr. Jean-Benoit Morin
Creator perspectiveRunning agent code on your own computer can consume all CPU, spawn unwanted processes, delete things it shouldn't, and expose environment variables; running on a server has similar problems with secrets and your environment.
AI Engineer · How I Learned to Stop Worrying and Love the Sandbox — Matt Brockman, E2B