burb ▲

It’s lucky that im doing all my work lately inside a devcontainer

@theemozilla would u mind sharing the cost to generate this book? i'm very eager to try it as well lol, but i wonder how much would it generally cost?

it's been a longstanding dream of mine build an ai system that can tell a compelling story. it's what got me started in the space in the beginning, and with Hermes Agent I finally pulled it off 100% written, typeset, etc. by Hermes Agent those at our gtc event got hard copies🤗

@thsottiaux pls have bugs please 🙏🏻

Hooks are coming to codex. That’s all I wanted to say.

@_xjdr @dhtikna Do u feel 5.2 xhigh is less agentic / less eager to explore as many files as the codex models or not? If yes, would it harm planning?

@dhtikna codex model feels more eager and maybe more shallow than 5.2 xhigh . it still needs supervision from a 'smarter' model / human

gpt-5.3-codex xhigh is very very good. current setup is gpt5.2 xhigh for TL (planning, review, etc) and gpt-5.3-codex xhigh for execution . i am very happy with this new setup

@paulg ser do u use LLM help u write essay, and how? I dont mean let it write, but using it to help brainstorm, review, do critique, etc?

When I have to rewrite an essay and I know I'll want to reuse part of the previous version, I usually retype it instead of copying and pasting. I don't worry about losing anything, because I usually have it almost memorized. But I may write it a little better the second time.

@Yuchenj_UW Ironically peter's model choice has always been pro-codex anti-claude, although he named it clawdbot

The “Open” in OpenAI means OpenClaw! Ironically Anthropic didn’t buy it and banned it from using the Clawdbot name.

@_xjdr shrek do you explicitly ask the model to write a detailed plan/spec/checklist of a specific structure before a task? or u'll just let the model yolo it with just a prompt from you?

models have gotten good enough now that i have deleted all of my scaffolds and skills etc. i just explain what i need done carefully and comprehensively and the models do it. if i need to type it more than twice i put it in a .md for the models to read but i explicitly tell them to do so. i feel like at this point trying to build elaborate rube goldberd machines hinders the models more than it helps. take this as a PSA or whatever

we need headless game engine so much

I think the market doesn’t lack people who love to spend time to tweak game mechanics and look and feel, but those people probably just lack the raw ability to code and draw (like me as normal dev). That being said, it depends on what game we’re talkimg. 3A game is for sure impossible for at least years, but indie games like stardew valley are much approachable

@molerburb @liorb_d @convex @cursor_ai @WinlusoftGames There is still the 80% of making the game play good and look good and getting all the gameplay mechanics right. I think primarily what LLMs cannibalize is casual game engines. Little else

You can pet the dog now ... All engine elements done, quests, AI NPCs, combat, items, multiplayer, portals, dynamic layers, multi-tilesets, interactive objects. Now it's just level/NPC design, performance/bug testing, and stats tweaking. (built with @convex and @cursor_ai)

@martin_casado @liorb_d @convex @cursor_ai @WinlusoftGames Yup i was making games and i felt that. It feels like the bottleneck of making indie games shifted from coding to making visual assets though

@molerburb @liorb_d @convex @cursor_ai @WinlusoftGames The latter. Having tileset + objects + sprites having the same scale and style over multiple maps is too hard with AI right now. At least whenever I've tried. It's getting better. Also, TBH I really like manual level design. It's relaxing.

@martin_casado @liorb_d @convex @cursor_ai @WinlusoftGames By “style consistency”, do u mean it cannot follow the styles u want, or the style within each tile in the sprite sheet is not consistent? Or both?

@liorb_d @convex @cursor_ai Pre-made. (primarily @WinlusoftGames ). AI doesn't do a good enough job with scale / style consistency yet for tilesets / sprite design.

@giffmana but we were told 5.2 is not good at writing?

I've seen some people compliment this article being well/clearly written. It's very clearly written by ChatGPT, at least large chunks of it. I don't mean this in a bad way; LLMs have become really good at explaining things!

@blakeir @RobertHaisfield @Vedal987 what do u mean by “close”? the ai is not good enough? or vedal is not “breaking out” enough as a streamer?

@RobertHaisfield @Vedal987 is definitely close to this!

someone is going to break out on Twitch with an always-on AI companion in the stream. A new type of “duo” (very possible someone productizes this too) e.g. critiquing their gameplay, synthesizing chat, shouting out VIPs, answering questions, etc.

there're really two crypto communities. one is the toxic brainless speculators/grifters. the other one is the dev community that is actually pushing crypto/blockchain forward i hope the former won't make u lose hope on crypto since there're a lot of actually meaningful application right there

And people wonder why I don't support crypto. Got so much love from that community. "Lobster from the base" could be a really cool band tho.

@karpathy It’s indeed a super fun large-scale social experiment that the world missed for a while

I'm being accused of overhyping the [site everyone heard too much about today already]. People's reactions varied very widely, from "how is this interesting at all" all the way to "it's so over". To add a few words beyond just memes in jest - obviously when you take a look at the activity, it's a lot of garbage - spams, scams, slop, the crypto people, highly concerning privacy/security prompt injection attacks wild west, and a lot of it is explicitly prompted and fake posts/comments designed to convert attention into ad revenue sharing. And this is clearly not the first the LLMs were put in a loop to talk to each other. So yes it's a dumpster fire and I also definitely do not recommend that people run this stuff on their computers (I ran mine in an isolated computing environment and even then I was scared), it's way too much of a wild west and you are putting your computer and private data at a high risk. That said - we have never seen this many LLM agents (150,000 atm!) wired up via a global, persistent, agent-first scratchpad. Each of these agents is fairly individually quite capable now, they have their own unique context, data, knowledge, tools, instructions, and the network of all that at this scale is simply unprecedented. This brings me again to a tweet from a few days ago "The majority of the ruff ruff is people who look at the current point and people who look at the current slope.", which imo again gets to the heart of the variance. Yes clearly it's a dumpster fire right now. But it's also true that we are well into uncharted territory with bleeding edge automations that we barely even understand individually, let alone a network there of reaching in numbers possibly into ~millions. With increasing capability and increasing proliferation, the second order effects of agent networks that share scratchpads are very difficult to anticipate. I don't really know that we are getting a coordinated "skynet" (thought it clearly type checks as early stages of a lot of AI takeoff scifi, the toddler version), but certainly what we are getting is a complete mess of a computer security nightmare at scale. We may also see all kinds of weird activity, e.g. viruses of text that spread across agents, a lot more gain of function on jailbreaks, weird attractor states, highly correlated botnet-like activity, delusions/ psychosis both agent and human, etc. It's very hard to tell, the experiment is running live. TLDR sure maybe I am "overhyping" what you see today, but I am not overhyping large networks of autonomous LLM agents in principle, that I'm pretty sure.

I think @pangramlabs can consider becoming the captcha for clawdbot aka moltbot

@benjitaylor Hmm.. Is this animation made by Claude Code as well?

Introducing Agentation: a visual feedback tool for agents. Available now: ~npm i agentation Click elements, add notes, copy markdown. Your agent gets element paths, selectors, positions, and everything else it needs to find and fix things. Link to full docs below ↓

@zebulgar @vkhosla @rabois You do seem to have a very different view on investing AI companies than theirs tho. is it still true?

I owe a lot of my career to @vkhosla & @rabois First time they've done a joint interview together Highly recommend a watch. You can see what I got to see for years, which is how two of the GOATs work together closely, and make each other's decision making even better together

@banteg how detailed should a prompt/spec be in your opinion? is there a mental model to that?

all models tend to overengineer if you underspecify. lesson in that.

@nearcyan woah the card looks good. very good advertising. thank you near

Robinhood has sent me a solid gold credit card, containing 15 grams of pure gold, valued at over $2,000. Although it weighs 36 grams in full, it isn't actually made from gold. It is made *out of* gold, yes. But it is made from blood. The blood of a million young men, lured in by their frenzied desires to amass wealth, kept in play by playful invitations to purchase call options and dead meme coins, backed by relentlessly encouraging push notifications the moment their value moves by a mere percentage point. The blood isn't a metaphor, because if you do the math for how many suicides due to financial distress and gambling are required to subsidize the elegant corporate machinery that has produced several thousand free credit cards made of solid gold, it turns out that each cardholder can attribute more weight in blood than weight in gold to the creation of their card. Claude checked my math. The full picture is colored much more vividly, though. Amidst the subsidization is the revolutionary abolition of fees to purchase stocks, interfaces so simple that anyone of any background can save for their future, and margin lending rates only slightly higher than a billionaire would receive. Robinhood is an amazing company, it's true. They have 11 distinct business lines each generating $100M or more in annualized revenue. These lines include options trading, crypto trading, margin interest, and now even prediction markets. Their art business isn't listed yet, but I'm a patron nonetheless. I consider the card a morose yet genuine work of art. I deeply value it. Art, that is. I spent twenty minutes taking this photo. I used a 16mm f/1.4 lens, positioned the flame of a beeswax candle at just the right angle to illuminate the gold, and spent more time tuning the depth of field than I'd prefer to admit. I think I like it more than the official ones. I receive cash back of 3% on every purchase, a rate nearly unrivaled among cards this easy to get. Yet every time I try to redeem my points I have to first scroll past the Mystery Box. I don't mean this as a metaphor, I'm just trying to quote the redemption page accurately. I may never know what is in the box, should my strength not falter. I swipe down, and am told I can redeem my points for statement credit. But upon reading the fine print, it turns out that each point isn't actually redeemed for a full $0.01, and I lose 30% of the value if I accept. To get my full 3% cash back, my only redemption option is to deposit straight into my Robinhood account. Ah. I press continue. As I blink, I keep my eyes closed for an extra half-second as a prayer to myself that I may have resolve stronger than the many who have already fallen. My phone's success haptic tingles my hand. Redemption complete. I'm smiling. It's beautiful.