Craig Allen retweeted
The statistics are even worse for studios that aren’t acquired.
Study: Nearly 61% of Acquired Game Studios Have Declined or Shut Down
A new analysis by Sebastian Cardoso, a former Riot Games, Crytek, and EA employee, examined 303 game studios acquired by major companies since 2000.
According to the findings, 60.8% experienced declining game performance or were shut down, with approximately one-third closing entirely.
The percentage of studios that declined or closed under each publisher:
Activision Blizzard: 100%
Microsoft: 63%
Sony: 58%
Tencent: 50%
Embracer Group: 43%
However, the findings depend on the study’s methodology and do not prove that acquisitions directly caused these outcomes.
Craig Allen retweeted
running a company nowadays feels kinda dystopian
my team is 50+ people in 12 countries, but i talk to my agents more than my human teammates
- voice note to grokbot
- mattpocock/skills /grill-me so the agent interrogates the idea before anything gets built
- spec drops into linear through the linear mcp
- exa + firecrawl agent reads every competitor site, pricing page and changelog
- ideabrowser.com to find ideas and trends
- apify pulls g2, reddit and app store reviews
- prompt: "find the 10 complaints that show up most and quote the users"
- context7 so the agent reads the latest docs before it touches any api
the design
- claude code + paper mcp
- 4 versions of every screen before a single line of code
- pick pieces from 2, it merges them. diverge then converge
- frontend-design skill from anthropic/skills + our own design system skill
- cleanshot screenshots of any ui i like go straight in with "make it feel like this"
the build
- pingdotgg/t3code as the control surface for every agent, from my phone too
- stablyai/orca runs a fleet of agents in parallel, each in its own git worktree
- graphify maps the whole codebase into a knowledge graph so agents navigate instead of guess
- headroom compresses tool outputs before they hit the model, so tokens go 20-95% further
- lsp plugins so the agent sees types and errors like an ide does
- security guidance hook flags risky edits before they land
- supabase mcp for the database, stripe for billing
- github mcp opens the pr, vercel spins up a preview
- loops run overnight, i check what shipped over coffee
the user testing
- lightpanda as the headless browser, way lighter than chromium so you can run hundreds at once
- agents click every flow on the preview like a first-time user
- one persona is impatient, one is confused, one is on a phone
- playwright + chrome devtools mcp for regression tests
the bugs
- sentry mcp catches the error
- agent reads the stack trace and opens a fix pr in under 10 minutes
- human just approves
the data
- posthog mcp cron at 7am on yesterday's funnel
- any step that drops more than 10%, the agent watches 20 session replays and tells me why
- stripe mcp pulls every cancellation reason into one doc every monday
the agents' tools
- orgo so every agent gets its own computer
- agentmail so every agent gets its own inbox
- agentphone so i can call them from the car
- composio connect-apps for every other connector
- akitaonrails/ai-memory, a git-backed markdown wiki so they remember what we decided last week
the skills
- wshobson/agents so skills work across claude code, codex and opencode
- my writing voice
- our design system
- how we write specs
- skill creator so the agents write new skills themselves
some helpful tricks
- one agent per platform. the x agent never touches email
- cheap model for triage, opus for anything that ships
- every output gets scored 1-10 by an llm judge. under 8 goes back
- every agent writes a 5-line log at night so i can read the whole company in 2 minutes
- every agent mistake becomes a new skill, so it only happens once
the team
- humans get version 5, not my voice note which they are happy about lol
- their job is creative feedback, relationships and telling the agents what they got wrong
a human pm reviews the funnel once a week. this reviews it every morning
running a company is now voice notes and inference
Craig Allen retweeted
Je vais dire quelque chose qui risque de déplaire à beaucoup de monde, mais tant pis.
Je crois qu’une partie de notre société est en train de perdre quelque chose de bien plus grave que sa compétitivité économique.
Elle est en train de perdre sa capacité à rêver.
Hier, Tesla a présenté une démonstration spectaculaire autour d’une voiture capable de quitter le sol. Un prototype, certes, pas encore la voiture volante de Retour vers le futur.
Mais quand même, bordel. On parle de technologies qui semblaient relever de la science-fiction quand nous étions gamins.
Et ce matin, en regardant certaines réactions françaises, qu’est-ce que je découvre ?
« Ça sert à rien. »
« Musk est un raciste. »
« Ça ne règle pas les problèmes de recharge. »
« Ça n’efface pas ses saluts nazis. »
Mais sérieusement, qu’est-ce qui nous est arrivé ?
À quel moment avons-nous perdu cette âme d’enfant qui nous faisait ouvrir grand les yeux devant une invention ?
Et le plus triste dans tout ça, c’est qu’on semble avoir oublié notre propre histoire.
La France, c’est le pays des frères Lumière, qui ont contribué à faire naître le cinéma.
C’est le pays de Louis Pasteur, dont les travaux ont révolutionné la microbiologie et la vaccination.
C’est le pays de Nicéphore Niépce, pionnier de la photographie.
C’est le pays des frères Montgolfier, qui ont envoyé l’être humain dans les airs à bord d’une montgolfière.
C’est le pays de Blaise Pascal, qui concevait déjà une machine à calculer au XVIIe siècle.
C’est aussi le pays du TGV, et l’un des deux pays qui ont donné naissance au Concorde.
Vous imaginez ce que représentaient ces inventions à leur époque ?
Des humains qui osaient accomplir ce que beaucoup pensaient impossible.
Des gens qui n’acceptaient pas les limites de leur temps.
Des visionnaires, des ingénieurs, des scientifiques, des entrepreneurs qui voulaient repousser les frontières du réel.
Nous avons été ce pays-là.
Alors comment avons-nous pu en arriver à un point où notre premier réflexe devant l’extraordinaire semble parfois être de chercher pourquoi il faudrait le mépriser ?
SpaceX récupère des fusées qui reviennent se poser sur Terre ? On parle immédiatement d’Elon Musk et de ses provocations.
Des robots humanoïdes commencent à réaliser des tâches physiques de plus en plus complexes ? « Ça va détruire des emplois. »
L’IA ouvre des perspectives gigantesques pour la médecine, la science, l’éducation et la création ? « Ça pollue, ça vole le travail des artistes, ça sert à tricher. »
Un entrepreneur annonce vouloir réaliser quelque chose d’incroyable ? « Encore un milliardaire qui veut s’en mettre plein les poches. »
Mais est-ce qu’on réalise seulement ce qui est en train de se passer ?
Nous vivons une époque où des fusées réutilisables se posent après leur mission, où l’IA accélère certaines découvertes scientifiques, où des robots apprennent à évoluer dans notre monde, où des personnes seules commencent à créer des choses qui nécessitaient auparavant des équipes entières.
Et devant tout ça, on serait devenus incapables de dire simplement : « Putain, c’est incroyable. Comment est-ce qu’ils ont réussi ça ? »
Je ne demande pas qu’on admire aveuglément les milliardaires. Ni qu’on ferme les yeux sur les conséquences sociales, environnementales ou politiques de leurs projets.
On peut condamner les propos d’Elon Musk tout en reconnaissant les prouesses des ingénieurs de SpaceX et de Tesla.
On peut défendre les travailleurs tout en trouvant la robotique fascinante.
On peut vouloir encadrer l’IA tout en étant émerveillé par ses possibilités.
L’esprit critique n’a jamais obligé personne à devenir cynique.
Et c’est justement ce cynisme qui me désespère.
J’ai parfois l’impression qu’on a remplacé l’admiration par la jalousie, le goût du risque par la peur de l’échec, l’ambition par la suspicion et le rêve par le sarcasme.
Craig Allen retweeted
We are now at "the fuck" levels.
"Our initial evaluation suggested that GPT-6-Astra's score was in the late eighties followed closely by Opus 5.5. But, as we started to investigate the mistakes that the models might have made, the models, especially Astra, insisted that they were right and that it was the expert-supplied answers that were erroneous. When we went back to the experts it turned out, in multiple instances, that that the experts found some subtle errors buried deep inside their calculations. As they revised their answers, Astra's score started to inch up and it finally ended up with 98% (49/50)."
Craig Allen retweeted
Character[.]ai financials:
- $13.8m on compute a month, then $10m a month
- bringing in $2.8m in subscriptions, sometimes less than $1m
- revenue from all sources $5.3m a month
Its bankruptcy shows the economics of AI don't work: 404media.co/character-ais-ba…
Craig Allen retweeted
When Spirit Airlines went bankrupt this year, Google paid $10 million for about 600 million of its employees' work emails and Teams chats. The people who wrote those messages didn't get a dime.
Hold onto that, because it explains why the startup in this post is offering its VP of semiconductors up to $9 million a year.
Grepped is a data broker. A company hands over its internal records, like its Salesforce data and accounting books, and Grepped cleans it all up and sells it to AI labs. Erik says he was at a $4 million monthly run rate about two weeks after he started.
The math on this job post is kind of wild. The hours are 9am to 9pm, seven days a week, so 84 hours. At the top of the range the VP makes roughly $2,000 an hour, and the founding scientist on the same list makes about $70.
Why would anyone pay a dealmaker 10 to 30 times what they pay a scientist? Look at the question he wants applicants to answer. "What does humanity know that it doesn't know it knows?"
That's basically his whole pitch in one sentence. AI labs already trained on the entire public internet. What they haven't seen is sitting inside companies, like the Slack thread where a senior engineer explained why the last fix broke production, or the ledger where an accountant made a judgment call and never wrote down why. People did that work for decades without ever thinking of it as knowledge. For a model it's gold, because it shows how experts actually make decisions when it counts.
Spirit put a price on it. Google beat Mercor's $7.5 million bid, and $10 million spread across 600 million messages comes out to less than 2 cents each. The money went to Spirit's creditors. The flight attendants' union went to court over the sale and pointed out the airline was selling their messages while it still owed them vacation pay.
Chips are where this gets really expensive. A chip company's design history might be the most protected IP in the world. You can't scrape it off the web, and the expert marketplaces paying engineers $85 to $200 an hour can only rebuild little pieces of it. Somebody has to walk into a chip company and talk them into selling 20 years of design files and failed chips. If an airline's chat logs went for $10 million, that archive goes for a lot more, and a broker only needs its cut on a couple of those deals to cover a $9 million salary.
So if you've spent the last ten years typing into Slack at your job, you've been building one of the most valuable datasets in AI, and whoever gets it sold is the one making $2,000 an hour.
we are hiring!
9am to 9pm, 7 days a week, in SF.
vice president, semiconductors: $4m–9m a year + equity
neural data lead (bci): $1.5m a year + equity
chip design data lead (semiconductors): $1.5m a year + equity
founding scientist, autonomous lab (sf): $300k–400k + equity
member of technical staff (sf): $340k–450k + equity
to apply, answer:
"what does humanity know that it doesn't know it knows?"
send your answer, resume + things your proud of to [email protected]
Craig Allen retweeted
今天和朋友聊了很久,关于 AI 对游戏、影视以及整个内容行业的影响。越聊越觉得,我们现在正站在一个很有意思的历史节点上。
聊天内容过长,我已经重新整理,补充部分论证例子,讨论如下:
01|我们正在经历一场被压缩到 1–2 年的“雅达利大崩溃”。
1983 年的北美游戏市场崩溃,原因非常复杂,但其中一个重要因素,就是大量低质量产品涌入市场,消费者逐渐失去信任,零售渠道也难以消化过剩的游戏。
今天 AI 带来的变化,某种程度上与之相似。过去,制作一款游戏、一部动画,至少需要相当的技术、资金与团队门槛。现在,这些门槛正在被快速降低。
Steam 上独立游戏的数量早已非常庞大;AI 辅助开发又进一步降低了程序、美术、音效甚至游戏设计的实现成本。未来,一个人制作过去需要十几人完成的产品,并不是不可想象的事情。
但问题是:生产能力的爆发,并不意味着消费能力也会同步增长。用户每天依然只有 24 小时。当所有人都能够生产内容时,内容本身就不再是稀缺品。我们也许会看到大量新项目出现,大量项目快速死亡,以及一轮又一轮的注意力争夺。
02|AI 不会立刻摧毁传统内容公司,但会极大压缩它们的生存空间。
雅达利大崩溃之后,并不是电子游戏消失了,而是整个产业重新洗牌,任天堂 FC 等新体系逐渐建立起来。值得注意的是,FC 的成功不仅来自硬件和游戏,更来自平台对内容品质、发行和渠道的重新组织。
今天的 AI 时代也可能如此。传统大型团队、资本密集型制作模式,依然有技术、管理、品牌与发行优势。但与此同时,过去需要数年、数千万成本才能完成的某些工作,未来可能被压缩到几个月甚至几周。
这会产生一个很残酷的结果:资本投入不再天然构成竞争壁垒,甚至可能成为沉重的历史包袱。当然,低成本不等于好产品。技术门槛下降后,审美、创意、决策与产品能力反而会成为更加直接的竞争因素。
03|我不认为 AI 时代的内容市场,会简单变成“谁便宜谁赢”。
看看今天的短剧市场,尤其是红果这样的免费短剧平台。它们带来的变化,不仅是制作成本和发行模式的变化,更重要的是重新组织了内容消费的方式。用户通过免费内容进入平台,平台通过广告等方式变现,而内容创作者则需要在这一套新的分发与商业体系中寻找自己的位置。
这提醒了一个问题:未来,真正掌握商业价值的,究竟是生产内容的人,还是掌握用户入口、分发机制与商业转化的平台?当内容越来越容易被生产时,流量并不会自动转化为创作者的利润。甚至有可能出现内容供给越来越丰富,而单个创作者的议价能力却越来越弱的局面。
04|AI 时代,低成本是一种生存优势,但流量不是最终目的。
我最近看到不少 AI 辅助开发的独立游戏 Demo,其中有些已经有相当不错的完成度。我认为,这类项目现阶段并不一定要急于证明自己能够成为下一款百万销量的游戏。它们完全可以先以低成本方式持续迭代,作为产品 Demo 吸引注意力、获取用户反馈,逐渐建立自己的核心受众。如果能够获得持续流量,相比过去需要投入大量营销预算,这本身就已经构成了一种成本优势。
但这里还有一个非常关键的问题:流量必须具有持续性,而且不能轻易被替代。靠一个爆款视频、一段 AI 视觉奇观获取的注意力,很容易被下一个更新鲜的作品夺走。而一款真正有趣的游戏、一个让人牵挂的角色、一套能够持续扩展的世界观,才有机会让用户主动回来。所以我越来越觉得,创作者应该从追求一次性传播,逐渐转向建立长期的用户关系。
05|我们真正需要警惕的是:创作成本下降了,但创作者的生存问题并没有被解决。
假设未来一个独立团队只需要 3 个人,就能够完成过去 30 人的工作。看上去成本降低了 90%。但如果市场上的同类产品数量增加了 100 倍,获客难度不断提高,那么这个团队真的比过去更容易生存吗?
不一定。甚至,AI 越普及,用户的注意力越昂贵。这可能让创作者陷入一个新的循环:制作越来越快,竞争越来越激烈,作品的商业生命周期越来越短,不得不持续生产内容来维持曝光。在这种情况下,低成本只能解决“如何活下去”的一部分问题,却无法自动回答“为什么用户要选择你”。
换个比喻,现在的内容创作,堪比在川渝新开一家火锅店,并存活3年以上。
06|我更期待 AI 时代的下一台 FC,而不是一万台更便宜的雅达利。
这里说的 FC,未必是某一台硬件。它可能是新的内容平台、新的游戏交互方式、AI 原生的娱乐形态,甚至是今天我们还无法定义的东西。下一代内容未必只是更便宜的电影、更快速生产的动画,或者更容易开发的游戏。正如早期电子游戏并不是简单复制电影,未来的 AI 原生内容,可能也不会停留在模仿今天的影视与游戏。
我自己非常期待这样的变化。但在那个节点到来之前,我认为内容创作者可能要做好一个现实准备:
深挖洞,广积粮。
尽可能降低成本,维持创作自由,积累自己的作品、技术、审美体系与核心用户。不要因为 AI 让生产变得容易,就误以为建立一个能够长期生存的创作事业也会变得容易。未来真正值得关注的,不是谁能生产最多内容,而是谁能够在无限内容的时代,让用户仍然愿意为他的作品付出时间。这可能才是 AI 时代内容创作者真正的护城河。
Craig Allen retweeted
Chamath says IP and software were "effectively rendered worthless" this week, after people decompiled Adobe's core products and published open-source versions, as every service goes headless
"The more important point I want to make though, is I sent this to Nick, there's been a couple things that happened in the last couple days, which I think are incredibly important."
"There have been people that decompiled the entire, the entire big five. What is the big five? Adobe has five very, very important products in their product suite. Freeberg may know them all, but I just know Photoshop. There's like a bunch of others. Okay."
"And these folks decompiled it, distilled it, and published open source versions of these things. And Rust, yeah. And I was so shocked, because I was like, hold on a second, these are like profound products that have existed and have had moats for years, and now they're just open source versions."
"And then Nikita was, I think, trolling, and he was building on that idea, and just talking about GTA, because it's about to get released. I think GTA 6 is about to come out, and we'll reconstruct this."
"Why is this important?"
"I think that IP and software was effectively rendered worthless two days ago."
"Because if you think about it, when everything is headless, so when Elon makes his services headless, Benioff made Salesforce headless a few months earlier, when all of these things are headless, and you can wrap them in agent tools, and you can connect them via MCP to other experiences, the value of IP does no longer exist in software. It doesn't exist."
"And I think that probably is the most profound implication."
"So, like, even in our business, we would argue with some of our customers all the time, because we'll say, oh, we'll build this for you, we'll own the IP. They'll say, no, we want the IP. We don't know how to negotiate it. And two days ago, I walked into the 8090 guys, I said, who cares? Who cares? But before, it was always a negotiation and a set of trade-offs. Now it's a who-cares thing, where, like, the IP is irrelevant."
"It still takes a while for the machinery of a business, and frankly the market, to understand. But that is the most important thing that happened this week in AI, in my opinion."
David Sacks says Anthropic is run by a "religious cult" and is doing the very things it says it fears most, like opening a wet lab while warning of bio risk, and that there's an argument for sending in a federal SWAT team
"But I do think that there are potential externalities from having the number one frontier model company run by a religious cult."
"So, for example, they are opening a Claude-powered wet lab in the middle of San Francisco. But at the same time, they're warning of bio risk. At the same time, they say that we have to pace the frontier, they're extending the frontier. At the same time that they are saying that there's a risk of superintelligence growing beyond our control, they're programming it to grow beyond our control."
"I think there are potentially huge externalities to the way they approach these things. And it's, like, insane that they call these decisions "safety." These are not safe decisions. These are crazy decisions, it seems to me."
"This is why I've called this a schizophrenic company, is they appear to be not only hypocritical, but they're doing the things that they say they are most afraid of."
"Like, what is going on here? This is bizarre. And I don't know where this all leads. I do think there's an argument for a federal SWAT team to come blasting in there."
Readers added context they thought people might want to know
The referenced open-source Adobe alternatives were not created via decompilation. Their developer describes them as clean-room reimplementations built from public specs, observed behavior and AI assistance in Rust.
github.com/storytold/phot…
arstechnica.com/ai/2026/10/sof…
frontiernow.dev/en/resources/a…
Craig Allen retweeted
Having long been aware of those tech bros who read my work as utopian porn, where they’ve taken that doesn’t exactly surprise me.
Craig Allen retweeted
UPDATE: AIs have achieved the highest *possible* score on the Mensa Norway IQ test - 151
3 years ago: cognitively impaired human (64 IQ)
2 years ago: average human
1 year ago: genius human
Today: literally off the charts
Next year?
Craig Allen retweeted
I wanted to quickly create a short video with Opus 5.5 about the history of AI, from "Attention is all you need" to AGI.
No stock footage, no image or video generators: every frame is rendered from code.
This 3-minute film was made 100% in code by Claude in Claude Code: ~7,400 lines of React/TypeScript (Remotion), every image drawn in SVG and Canvas, an open-source TTS voice, and a score synthesized in Python. No stock footage. No image generators.
Id say pretty cool! Tookl about 1h and 7% weekly rates.
Craig Allen retweeted
Elon Musk y su empresa Neuralink han conseguido leer la mente. Es un hito histórico que cambiará la vida de mucha gente.
Piensas algo y la máquina es capaz de leerlo y reproducirlo. Aquí un hombre que le vuelve a decir “te quiero” a su esposa después de muchos años mudo. ❤️
Craig Allen retweeted
big AI news
Google just demonstrated a recursive self improvement loop for AI discovery
Google/DeepMind researchers introduced Dream-RSI, a system where an AI agent improves how it explores problems by replaying its past discovery attempts, testing thousands of alternative strategies cheaply, then deploying the better strategy in the next round.
Across algorithm design, mathematical optimization, and GPU kernel engineering, it matched or improved discovery quality while cutting search costs dramatically, in one setting reducing agent calls by up to 162x. 👀
Importantly, it improves the exploration policy, not the underlying model weights.
Craig Allen retweeted
Very funny that xAI was created to stop Anthropic which was created to stop OpenAI which was created to stop Google which would've actually never developed ASI because it was afraid of cannibalizing its ads revenue from search.
Craig Allen retweeted
Stanford built an AI that predicts when you'll die from a single night of sleep.
And the results are terrifyingly accurate.
It’s called SleepFM. It reads your brain, heart, and breathing signals overnight and predicts 130 future diseases before symptoms appear.
The model was trained on nearly 600,000 hours of clinical sleep data from 65,000 people.
It tracks everything simultaneously: brain waves, heart rhythms, breathing patterns, eye movements, and muscle activity.
By using a clever training technique called leave-one-out contrastive learning, the AI learned the deep "language of sleep."
It mapped out how different body systems interact when you are unconscious.
The accuracy numbers are staggering:
• Overall Mortality Risk: 84% accuracy
• Parkinson’s Disease: 89% accuracy
• Dementia: 85% accuracy
• Heart Attacks & Heart Failure: 80-81% accuracy
• Cancers: High precision mapping across the board
But the craziest part?
The model’s greatest predictive power didn't come from a single bad metric.
It came from spotting physiological discordance.
That’s what happens when parts of your body are completely out of sync while you sleep, like a brain showing deep sleep patterns while your heart rhythm looks entirely alert.
Preventative medicine just entered a completely different dimension.
We spent decades treating sleep like a passive state of rest.
Turns out, it’s a living trailer for your future health.






































