David Koh

AI, Software

@Berkeley_EECS @Microsoft

[David Koh @dooytr]

@jeremybernier I had an interview where it was return me the avg size of all images on Wikipedia. It was a coding question. Said I could use ai or not. Didn’t really get what the interviewer was grading

[David Koh @dooytr]

@PositiveGrid ‘s support is so bad. It’s all generic AI responses that I can ask ChatGPT. No human support. I don’t know if I can trust buying another product from them in the future with poor support

[Amanda Orson @amandaorson]

The faster technology moves, the more I think about Bezos' question

What won't change in the next 10 years?

Things I've been writing down over time:

- Humans will always need shelter, food, energy, and healthcare.
- The desire for ownership and the accumulation of wealth.
- The physical world will move more slowly than the digital one.
- Every increase in technological capability, especially AI, will require more energy.
- People and businesses will continue to need access to capital.
- Capital will continue to seek returns that exceed inflation.
- Underwriting methods evolve, but demand for credit (loans) is persistent.
- Trust remains scarce and becomes increasingly valuable as content, code, and fraud become cheaper.
- Verified identities and reputation becomes more important as information becomes abundant and synthetic.
- Long-term wealth creation and dynastic (multi-generational) thinking predate modern technology, and will persist.
- Coordination and transaction costs never fully disappear; market friction will continue to justify the existence of firms and intermediaries.
- People will continue to compete for status.
- Consumers will pay a premium for products and services that confer status.
- Time remains fixed at 24 hours per day.
- But attention is a finite resource and an enduring constraint.
- Products that credibly save time (or enable delegation) have a perpetual market.
- Inaccessible, proprietary data will be a persistent moat. The more inaccessible and difficult to aggregate, the deeper the moat.
- People want accountability, recourse, and clearly identifiable responsibility when things go wrong.
- Regulation consistently lags technological innovation.
- Compliance requirements, licensing, and regulatory moats persist even when machines can perform the underlying task.
- Local knowledge remains valuable and difficult to replicate.
- Heterogeneous markets (like real estate) continue to reward people with deep contextual understanding.
- Incumbent organizations tend to underinvest in disrupting their own businesses, which always creates opportunities for challengers.

Bezos' insight on what wouldn't change in 10 years was "Customers will always want lower prices and faster delivery."

It's boring/ true, but I think that's the point.

Everything we build today can and will be rebuilt more cheaply, faster by someone else.

Build on the invariants, not the trends.

What have I missed?

[David Koh @dooytr]

I like @grok. I find it fun to use. However I get rate limited way too often on free tier. Probably after like 5 prompts. I never get rate limited on @GeminiApp on the chrome browser and it's become my default for quick casual conversations. (Unrelated to agentic coding)

[David Koh @dooytr]

@lufthansa @aixarizzo Had a bad experience too but @lufthansa doesn't compensate and don't respond in their links

[David Koh @dooytr]

@neondatabase Building a surgical db failover platform. Where we can failover impacted dbs at the right granularity and reduce blast radius.

[David Koh @dooytr]

@sama Is it just companies or for individuals too?

[X Freeze @XFreeze]

Elon Musk: "If you punish people too much for failure, then they will respond accordingly, and the innovation you will get will be very incrementalist

Nobody's gonna try anything bold for fear of getting fired or being punished in some way. So risk-reward must be balanced and favor taking bold moves, otherwise it will not happen"

[🍂 @Lovandfear]

“To be happy you must eliminate two things: The fear of a bad future and The memory of a bad past.”

- Seneca

[David Koh @dooytr]

@Polymarket @legorobotdude lol "podslop"

[David Koh @dooytr]

@Polymarket That never works

[David Koh @dooytr]

"AI coding agents" aren't models. They're runtimes.

The model proposes. The harness decides whether it actually happens — permissions, tool validation, context budgets, memory, sub-agents, cancellation, cleanup.

https://t.co/NsfK3xjUnU

[Brivael Le Pogam @brivael]

Elon Musk avait dit un truc qui m'avait marqué sur l'allocation de ressources. En substance : passé un certain niveau de richesse, l'argent n'est plus de la consommation, c'est de l'allocation de capital.

Cette phrase change tout.

L'économie, dans le fond, c'est juste un problème d'allocation. Tu as des ressources finies et des usages infinis. Qui décide où va quoi ?

Imagine une cour de récré. 100 enfants, des paquets de cartes Pokémon distribués au hasard. Tu laisses faire. Très vite, un ordre émerge. Les bons joueurs accumulent les cartes rares, les collectionneurs trient, les négociateurs trouvent des deals. Personne n'a planifié. Et pourtant chaque carte finit dans les mains de celui qui en tire le plus de valeur. Le système maximise le bonheur total de la cour. C'est ça, la main invisible.

Maintenant fais entrer la maîtresse. Elle trouve ça injuste. Léo a 50 cartes, Tom en a 3. Elle confisque, redistribue, impose l'égalité. Trois effets immédiats. Les bons joueurs arrêtent de jouer, à quoi bon. Les mauvais n'ont plus de raison de progresser, ils auront leur part. Les échanges s'effondrent. La cour est égale, et morte. Elle a maximisé l'égalité, elle a détruit le bonheur.

Le problème de la maîtresse, c'est qu'elle ne peut pas avoir l'information que la cour avait collectivement. C'est le problème du calcul économique de Mises, formulé en 1920. L'URSS a essayé de le résoudre pendant 70 ans avec le Gosplan. Résultat : pénuries, queues, effondrement. Pas parce que les Soviétiques étaient bêtes, parce que le problème est mathématiquement insoluble en mode centralisé.

Quand Musk a 200 milliards, il ne les consomme pas, il les alloue. SpaceX, Starlink, Neuralink, xAI. Chaque dollar est un pari sur le futur. Et lui a un track record. PayPal, Tesla, SpaceX. Il a démontré qu'il sait identifier des problèmes immenses et y allouer des ressources avec un rendement spectaculaire.

L'État aussi a un track record. Hôpitaux qui s'effondrent, éducation qui décline, dette qui explose, services publics qui se dégradent malgré des budgets en hausse constante. Le marché identifie les bons allocateurs, la politique identifie les bons communicants.

Le profit n'est pas une finalité, c'est un signal. Il dit : tu as alloué des ressources rares vers un usage que les gens valorisent suffisamment pour payer. Plus le profit est gros, plus la création de valeur est grande. Quand Starlink est rentable, ça veut dire que des millions de gens dans des zones rurales ont enfin internet. Quand un ministère est en déficit, ça veut dire qu'il consomme plus qu'il ne produit. L'un crée, l'autre détruit, et on appelle ça redistribution.

Dans nos sociétés il y a deux catégories d'acteurs. Les entrepreneurs et les bureaucrates. L'entrepreneur prend un risque personnel pour identifier un problème, mobiliser des ressources, créer une solution. S'il se trompe il perd. S'il a raison, ses clients gagnent, ses employés gagnent, ses fournisseurs gagnent, l'État collecte des impôts. Il est la cellule de base du progrès humain.

Le bureaucrate ne prend aucun risque personnel. Son salaire est garanti. Au mieux il maintient une rente existante. Au pire il la détruit par excès de réglementation, mauvaise allocation forcée, incitations perverses qui découragent ceux qui produisent. Mais dans aucun cas il ne crée.

Regarde les 50 dernières années. iPhone, internet civil, SpaceX, Tesla, Google, Amazon, Stripe, mRNA, ChatGPT. Toutes des inventions privées, portées par des entrepreneurs, financées par du capital risque. Pas un seul ministère n'a inventé quoi que ce soit qui ait changé ta vie au quotidien.

La France est devenue le laboratoire mondial de la dérive bureaucratique. 57% du PIB en dépenses publiques, record absolu. Une administration tentaculaire, une fiscalité qui pénalise la création de richesse. Résultat : décrochage face aux États-Unis, à l'Allemagne, à la Suisse. Fuite des cerveaux. Désindustrialisation. Dette qui explose.

Et le pire c'est que la mauvaise allocation s'auto-renforce. Plus l'État prélève, moins les entrepreneurs créent. Moins ils créent, moins il y a de base fiscale. Plus l'État s'endette et taxe. Boucle de rétroaction négative parfaite. La maîtresse pense qu'elle aide, et chaque année la cour produit moins.

Dans nos sociétés, ce sont les entrepreneurs, toujours, qui font avancer la civilisation. Les bureaucrates au mieux maintiennent une rente, au pire la détruisent. Aucune société n'a jamais progressé en taxant ses créateurs pour subventionner ses gestionnaires.

La question n'est jamais qui a combien. C'est qui alloue le mieux la prochaine unité de ressource pour maximiser le futur de l'humanité. La réponse depuis 200 ans n'a jamais changé. Ce ne sont pas les fonctionnaires.

[Marc Randolph @marcrandolph]

You cannot install a culture of experimentation.

You can only model it.

Culture isn’t what’s written in the employee handbook or posted on the lobby wall.

Culture is what people observe — who gets celebrated, who gets promoted, who gets fired, and why.

Everything else is decoration.

[Lenny Rachitsky @lennysan]

My biggest takeaways from Claude Code's Head of Product @_catwu:

1. Anthropic’s product development timelines have gone from six months to one month, sometimes one week, sometimes one day. Part of this acceleration is access to the latest models (i.e. Mythos). Another is shipping new products into “research preview,” making clear it's early, experimental, and might not be supported forever. Another is an evergreen "launch room "where engineers post ready features and marketing turns around announcements the next day.

2. The PM role is shifting from coordinating multi-month roadmaps to enabling teams to ship daily. As Cat puts it, “There should be less emphasis on making sure you are aligning your multi-quarter roadmaps with your partner teams and more emphasis on, OK, how can we figure out the fastest way to get something out the door?”

3. The most efficient shipping unit is an engineer with great product taste. On Cat’s team, many engineers go end-to-end—from seeing user feedback on Twitter to shipping a product by the end of the week—without a PM involved. Also, almost all the PMs on the Claude Code team have either been engineers or ship code themselves, and the designers have been front-end engineers. The roles are merging, and the most valuable skill is product taste, not job title.

4. Build products that are on the edge of working. Claude Code’s code review product failed multiple times because earlier models weren’t accurate enough. But because the prototype was already built, they could swap in Opus 4.5 and 4.6 and immediately test whether the gap was closed. Teams that wait for the model to be ready will always be a cycle behind.

5. The most underrated skill for building AI products is asking the model to introspect on its own mistakes. Cat regularly asks the model why it made an unexpected decision. The model will explain that something in the system prompt was confusing, or that it delegated verification to a subagent that didn’t check its work. This reveals what misled the model so the team can fix the harness.

6. Every model release forces their team to revisit existing products and audit their system prompt to remove features the model no longer needs. Claude Code’s to-do list was a crutch for earlier models that couldn’t track their own work. With Opus 4, the model handles it natively. Features built as scaffolding for weaker models become debt when the model catches up—so the team actively strips them.

7. Anthropic employees build custom internal tools instead of buying SaaS products. A sales team member built a web app that pulls from Salesforce, Gong, and call notes to auto-customize pitch decks—work that used to take 20 to 30 minutes now takes seconds. Their core stack is Claude Code, Cowork, and Slack. No Notion, no Linear, no Figma.

8. People underestimate how much Claude’s personality contributes to its success. As Cat describes it, “When you reflect on everyone you’ve worked with, there’s just some people where you’re like, I really like their energy, their vibe.” Claude is designed to be low-ego, positive, competent, and earnest—qualities that make it feel like a great coworker, not just a tool. This isn’t cosmetic; it’s what makes people want to use Claude for hours every day. The team has a dedicated person, Amanda, who “molds Claude’s character,” and it’s one of the hardest roles at the company because success is so subjective.

9. The future of work is managing fleets of AI agents, not doing the work yourself. Cat sees a clear progression: first, individual tasks become successful. Then people start running multiple tasks at the same time (multi-Clauding). Next, people will run 50 or 100 tasks simultaneously, which will require new infrastructure—remote execution, better interfaces for managing tasks, agents that fully verify their work, and self-improving systems that incorporate feedback. The human role shifts from doing the work to knowing which tasks to look into, verifying outputs, and giving feedback that makes the system better over time.

10. Hire people who lean into chaos and face every challenge with a smile. At Anthropic, there are weeks when a P0 on Sunday becomes a P00 by Monday and a P000 by Monday afternoon. If you get too stressed about any one thing, you’ll burn out. Their team looks for people who can look at a hard challenge and say, “Wow, that’s gonna be hard. But I’m excited to tackle it and I’m gonna do the best that I possibly can.” This mindset—optimism, resilience, and comfort with constant change—is increasingly essential as the pace of AI development accelerates.

Don't miss the full conversation: https://t.co/1wOUHcdYQN

[David Koh @dooytr]

@jordymaui Hermes can also support multiple models.

[David Koh @dooytr]

AI may replace many jobs. But one job is safe from AI.

Politicians

[David Koh @dooytr]

Regression in output came from the harness not the model.

I mentioned in my latest substack post that the harness is more important than the model.

[ClaudeDevs @ClaudeDevs]

The issues stemmed from Claude Code and the Agent SDK harness, which also impacted Cowork since it runs on the SDK.

The models themselves didn't regress, and the Claude API was not affected.

[David Koh @dooytr]

Forget the laptop setup. I’ve been experimenting with Discord as a portable IDE. It’s the ultimate 'Code To Go' workflow for agentic AI.

https://t.co/cZz4cmJWGh

[David Koh @dooytr]

@garrytan What makes openclaw give you the time of your life? I like hermes