Principle · MoneyThe brain is the ultimate model of hardware-software co-design: hardware and software are one and the same and have co-evolved. When taking pieces of brain-inspired design onto other computing substrates or types of physics, one has to be very careful, because whatever is native to that form factor is probably the right way to improve it.
Consequential creator advice requires independent verification before Today can promote it.
Y Combinator · What If We Stopped Using GPUs? | YC Paper Club
Method · MoneyA method for managing prompts in production: keep multiple versions of prompts so evolution is visible, treat a versioned prompt as the trigger for eval automation, promote a previous version back into production when regressions appear, and use the API/CLI rather than manual UI clicks for promotion.
Method, not a verified protocol.
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Recommendation · MoneyAnnabell Schäfer advises that in regulated spaces like financial services, you should define hard limits (e.g., 'never give investment advice') and ingrain them in your evaluation set. Otherwise prohibited outputs could reach production unnoticed and you won't know when the prohibited behavior happens.
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Recommendation · BodyDo 10 sets of 10 calories or one set of 100 calories on the air bike every week or twice a week, done pretty intensely, and you will improve your conditioning without a doubt. You can get an excellent quality conditioning session in less than 10 minutes that makes genuine progress.
Garage Strength · How To Get Strong and Still Be Athletic w/ Sika Strength
RecommendationRogan urges people not to push further away from those who disagree but to move closer and talk to them. He also calls for stopping the antagonistic us-versus-them mentality and says he does not want to align with political parties because he does not think that is smart.
PowerfulJRE · Joe Rogan Experience #2391 - Duncan Trussell
Creator perspective · MoneyLincoln Murr describes three core agentic trading use cases that exist today: (1) fully managed portfolios where the user specifies an allocation (e.g. 80% S&P, 5% Bitcoin, 15% T-bills) with guardrails and lets the agent run; (2) autonomous execution of scenario-based trades predefined in advance; and (3) translating a high-level natural-language investment idea into actual financial primitives and a strategy.
speaker's framing of current use cases
AI Engineer · Why AI Agents Will Need Their Own Bank Accounts — Lincoln Murr, Coinbase
Report · BodyThe 100-cal or 50-cal air bike test is a great benchmark for field-sport conditioning. Around three minutes for 100 calories is good depending on the sport; around a minute for 50 calories means you're exceptionally well conditioned. Nobody reaches the three-minute mark on 100 calories without putting time into their conditioning — you can't fake it. Benchmarks differ by sport: a front-row rugby player won't have the same conditioning profile as a jiu-jitsu or judo athlete.
Garage Strength · How To Get Strong and Still Be Athletic w/ Sika Strength
Creator perspectiveJoe Rogan proposes what he sees as the only way to get everyone on your side: borders are closed, and you find out which illegal immigrants have committed felonies. If an illegal immigrant has committed X amount of felonies, they have to leave the country — if you've been robbing people for the last 10 years and you're an illegal, you have to leave the country, 'that makes sense.' But if you've been here for 25 years, have a family, kids go to school here, speak the language, and are a contributing member of the community, deporting them is crazy.
Rogan's endorsed proposal, clearly attributed.
PowerfulJRE · Joe Rogan Experience #2391 - Duncan Trussell
Creator perspectiveJoe Rogan argues that when people have been here for 20 years, mass deportation is crazy — 'come on. That's crazy.' He says if immigrants have been productive members of society for 20 years — no criminal record, worked the entire time, paid taxes — find them a pathway to citizenship. He also notes that shipping lower-income and middle-income people who are kind of getting by to a country where they've never lived is crazy to ask, and that deporting long-term contributing illegal residents who have families and kids in school shows no heart — you won't get any reasonable people to go along with it.
Rogan's endorsed position, clearly attributed.
PowerfulJRE · Joe Rogan Experience #2391 - Duncan Trussell
Report · MoneyGood and bad feedback to cells is defined through asynchronous and synchronous stimulation: if the cells do something bad in the game, they get asynchronous stimulation (firing at different times at random) to reinforce not doing that action; if they do something good, they get synchronous stimulation. Per the free energy principle from Karl Friston, asynchronous stimulation is something the brain wants to avoid. The team had to predict surprise: they use the critic in the silicon decoder to figure out how far the action is from the silicon critic and how surprising that specific action is, and feedback is scaled based on surprise — if the critic expected an action to be good and the chosen action was really bad, it gets negative feedback. On top of the surprise scaling, feedback is scaled
The free energy principle link is presented as the team's belief/interpretation.
Y Combinator · What If We Stopped Using GPUs? | YC Paper Club
Recommendation · MoneyLincoln Murr recommends giving an AI agent only a small amount of agency — the human still directs the agent — starting with small amounts of money for microtransactions before granting larger financial access, and granting access to an isolated portfolio only after guardrails, rules, and safety policies are in place.
creator recommendation, attributed to Lincoln Murr
AI Engineer · Why AI Agents Will Need Their Own Bank Accounts — Lincoln Murr, Coinbase
Method · MoneyA method for debugging production agents: define failure-mode categories captured in order of occurrence, filter traces by score conditions (e.g., dispute not resolved, security risk flags) instead of scanning hundreds of thousands of traces, and save filtered views for reuse.
Method, not a verified protocol.
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Recommendation · BodyBefore starting a running training block, check whether the person has trained enough to actually train running; 'training to train' is a distinct phase from a running block.
Garage Strength · How To Get Strong and Still Be Athletic w/ Sika Strength
RecommendationJoe Rogan encourages direct help: giving out water is not that hard, you can help people without filming it or doing it for any reason, and helping at soup kitchens will make you feel like a piece of shit by comparison if you're not doing it.
PowerfulJRE · Joe Rogan Experience #2391 - Duncan Trussell
Creator perspective · MoneyDoneyli argues that manually looking at traces is still a very important process despite its cost — Annabell spent days doing it — but that once a system is in production it is not feasible for a human to look at thousands of traces. The resolution is to curate a representative sample of production traces for a human expert to review.
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Creator perspective · MoneyBanks face financial liability, legal liability, and a bunch of risk when putting out LLM-powered customer-facing solutions because LLMs are inherently non-deterministic. When evaluating a customer-facing agentic solution, leadership or teams will ask how to evaluate it after the first prototype or demo is built. Reduce the unpredictability of LLM-powered agents so you can reproduce outcomes as much as possible and figure out failure modes.
described
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Principle · MoneyWhen testing whether the same model performs better with a different prompt, change only the prompt and not five other things. Changing multiple variables at once makes results hard to interpret; isolating concerns lets you see whether a change better handles your existing data.
AI Engineer · Evals for High-Stakes AI Agents — Doneyli De Jesus & Annabell Schäfer, Langfuse
Creator perspective · MoneyIf AI labs run their best models on basic cryptographic assumptions and fail to break them, that failure itself gives tremendous confidence the assumptions are valid. Even a partial AI result — e.g., shaving 10 bits of security off a problem — is valuable evidence the assumption is still worthwhile.
The a16z Show · Can AI Break Cryptography? What the Experts Actually Think
Creator perspective · MoneyThe types of cryptography threatened by advanced AI are the same types under threat from a quantum computer — a threat already being prepared for. If the AI threat arrives before the quantum computer threat, it doesn't make it new — mitigations are already deployed. People have already begun to treat elliptic curve cryptography as something that may be vulnerable soon and have deployed mitigations for those exact concerns. The quantum threat is really only about elliptic curve-type systems and RSA.
Conditional on AI attacks materializing at all.
The a16z Show · Can AI Break Cryptography? What the Experts Actually Think
Principle · BodyFor most athletes in season, the main decay is in absolute strength — the absolute strength number falls off more than anything else. Conditioning values will increase or at least be maintained, and speed and power numbers tend to improve throughout the season. There is also some fall-off in range of motion and strength numbers.
Garage Strength · How To Get Strong and Still Be Athletic w/ Sika Strength
