Musk vs OpenAI
AI Valley — aivalley@mail.beehiiv.com
Reçu le
mercredi 29 avril 2026 à 14:43
Source
AI Valley
Message-ID
20260429.144312.5141042a@mta31Version nettoyage
v1.0.0 (ok)Brut (HTML rendu, sandboxé, ressources externes bloquées)
Nettoyé (Markdown — clean déterministe)
~[Sign up](https://www.theaivalley.com/)~~ | ~~[Follow us on X](https://twitter.com/heyBarsee)~~ | ~~[Sponsor](mailto:barsee@aivalley.ai)~ View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/c167fb98-39eb-42d4-a114-56fe02753cac/2__1_-min__2_-min.png?t=1744033912) Follow image link: (https://www.theaivalley.com/subscribe) Caption: Together with View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/a305da31-7cc1-46d4-bf3d-e975b91dbcfa/6303a615ab1b900654aad3c9__1_.png?t=1777471432) Follow image link: (https://softr.io/build-with/ai-valley) Caption: ---------- howdy, it’s Barsee again. happy wednesday, AI family, and welcome back to AI Valley. **here are the biggest things worth knowing today:** * Musk sues OpenAI, calls Altman’s shift “looting” a charity * Researchers train an AI that thinks like it’s 1930 * Thousands of humanoid robots are entering real warehouses * Anthropic has now surpassed OpenAI in revenue * Plus trending AI tools, posts, and resources Let’s _dive_ into the Valley of AI… ---------- ---------- #### **SOFTR** ## [The AI app builder you can run a business on](https://softr.io/build-with/ai-valley) View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/7c6c72ad-c56c-4870-9c1b-899504095f95/AI_Ad__1_.png?t=1777471406) Follow image link: (https://softr.io/build-with/ai-valley) Caption: Courtesy: Softr Describe the business tool you need, and Softr AI builds the whole thing — database, logic, logins, permissions, security, and hosting included. Switch between AI and visual editing whenever you want (so no re-prompting just to fine-tune). Think client portals, internal tools, CRMs, and more. No code required. [Start today with 200 free AI credits.](https://softr.io/build-with/ai-valley) ~_*This is sponsored_~ ---------- ---------- #### **THROUGH THE VALLEY** ##### 1/ Musk sues OpenAI, calls Altman’s shift “looting” a charity View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/ef29ad61-04bd-4710-9db9-07d460ea04e2/Elon-Musk-and-Sam-Altman215-04282026-fbb4c904d2c545b4bd25346666224367__1_.png?t=1777470329) Caption: Elon Musk says he’s suing OpenAI over its shift from [nonprofit to for-profit](https://www.bloomberg.com/news/articles/2026-04-28/musk-lawyer-says-altman-made-mockery-of-openai-public-mission?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTc3NzQzNjcwMCwiZXhwIjoxNzc4MDQxNTAwLCJhcnRpY2xlSWQiOiJURTdSRTZLSzNOWTkwMCIsImJjb25uZWN0SWQiOiJCMzZENUE5QzIxMDQ0NjU4OTFBMTc1MTVDRDNBQkZFNiJ9.Vmp8keSnmoLgBiND9IghO3fieMbkedWND_2trW15z6w&leadSource=uverify+wall), arguing it turned a public mission into private gain. He claims his early funding and reputation helped build the company, only for it to abandon its original principles. Musk warned this could set a precedent for “looting every charity in America,” while OpenAI argues the lawsuit is just an attempt to weaken a competitor. **Why this matters:** This is really about who controls AI labs once big money is involved. If courts side with Musk, “mission-first” structures lose credibility fast. If not, investors are backing companies where control and ownership don’t fully align. ##### 2/ Researchers train an AI that thinks like it’s 1930 View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/110090db-a4e3-461f-81c5-9aa36b6303f5/chat-gpt-1930s-ai__1_.png?t=1777470375) Caption: Researchers trained a 13B model, [Talkie](https://talkie-lm.com/introducing-talkie), only on data from before 1931, using books, newspapers, and public domain records. To make it conversational, they used old etiquette guides and cookbooks. Despite never seeing modern code, it can still write working Python by adapting patterns. **Why this matters:** The model didn’t “know Python.” It likely saw older programming patterns (math notation, early code, structured logic) and learned how to combine them. So when asked for Python, it’s not recalling syntax, it’s guessing a modern format that fits those patterns. Sometimes that guess is good enough to run. But it’s not reliable, and it breaks more easily than models trained on real code. ##### 3/ Thousands of humanoid robots are entering real warehouses View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/c8eb82e5-573b-4a38-b972-cb8d7952a434/jpDBojorhmqyyUet__1_.png?t=1777470784) Caption: RobotEra is deploying humanoid robots in[ 10 logistics hubs](https://x.com/coinbureau/status/2049272449196499408?s=20), handling sorting and package movement in real operations. The company has raised $400M+ and is already running across 10+ facilities, with robots reaching ~85% of human efficiency while operating continuously. [Japan Airlines](https://x.com/Reuters/status/2049424702536405364?s=20) is also testing humanoids for baggage handling and aircraft cleaning at Haneda Airport starting in May, pointing to growing adoption driven by labor shortages and rising travel demand. **Why this matters:** At ~85% efficiency, robots don’t need to be perfect, just cheaper over time. Warehouses are built for humans, so humanoids can plug in without redesign. If this works here, it spreads fast to similar industries. ##### 4/ Anthropic has now surpassed OpenAI in revenue View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5be50a79-1d58-4cf9-98e0-3a20d924537a/image_-_2026-04-29T192522.948__1_.png?t=1777470875) Caption: Anthropic has reportedly reached [~$30B ARR](https://x.com/NoLimitGains/status/2049213286311436597?s=20), ahead of OpenAI’s ~$24B, after scaling from ~$1B in about 15 months. The growth is driven by enterprise, not consumers, with 1,000+ companies spending over $1M annually on Claude. OpenAI, meanwhile, serves a massive user base but carries higher infrastructure and training costs. **Why this matters:** This is shaping into two different AI business models. OpenAI is optimizing for reach, building a massive user layer. Anthropic is optimizing for revenue density, extracting more value per customer. If enterprise demand continues to dominate, depth of integration may matter more than distribution, which changes how the next wave of AI companies are built. ---------- ---------- #### **PRESENTED BY ADOBE** ## [Adobe just launched an AI creative assistant](https://firefly.adobe.com/) View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/a833e22e-fc38-4d2b-b8d4-3bf758b80601/Firefly_AI_Assistant_SocialAssets__1_.png?t=1777471105) Follow image link: (https://firefly.adobe.com/) Caption: Courtesy: Adobe Now live in the app, Adobe just turned Firefly into a true all-in-one creative AI studio with its new Firefly AI Assistant. It works like an agent. You describe the outcome, and it plans and executes multi-step workflows across apps like Photoshop, Premiere, Illustrator, and more, while you stay in control of the creative direction. Beyond that, Adobe upgraded Firefly’s editing stack. Video gets studio-quality sound, advanced color controls, and Adobe Stock integration, while images get more precise tools like Precision Flow and AI Markup. It also now supports 30+ top AI models, including Kling 3.0 (and Omni), Veo 3.1, Runway Gen-4.5, and ElevenLabs Multilingual v2, giving creators far more flexibility in how they create. [Adobe Firefly](https://firefly.adobe.com/) ---------- ---------- # TRENDING TOOLS -------------------- * [Remio](https://www.remio.ai/) >** **Set up once. Remio captures everything and gets the work done — docs, emails, slides, reports ^~(sponsored)~^ * [Lovable](https://lovable.dev/blog/mobile-app) > Build full-stack web apps straight from your phone using natural language * [Talkie](https://talkie-lm.com/introducing-talkie) > Researchers unveil a 13B “vintage” language model trained to mirror 1930s-era language and style * [Claude](https://www.anthropic.com/news/claude-for-creative-work) > Expands creative workflows with new connectors, including Adobe tools and Affinity by Canva ---------- ---------- # WHAT I'M CONSUMING -------------------- * [ElevenLabs launches 50 templates to turn its voice AI into your next employee](https://x.com/ElevenLabs/status/2048794727881789930) * [Study shows 35% of newly published websites since ChatGPT launch are AI-generated](https://www.404media.co/study-finds-a-third-of-new-websites-are-ai-generated) * [Researchers built an AI that thinks it is 1930](https://www.marktechpost.com/2026/04/27/meet-talkie-1930-a-13b-open-weight-llm-trained-on-pre-1931-english-text-for-historical-reasoning-and-generalization-research/) * [xAI is quietly building a social network for AI-generated art](https://www.testingcatalog.com/xai-rolling-out-custom-shareable-imagine-templates-for-grok) * [How ChatGPT serves ads to you](https://www.buchodi.com/how-chatgpt-serves-ads-heres-the-full-attribution-loop) * [Beijing gives Meta weeks to unwind its latest AI acquisition](https://www.wsj.com/tech/ai/meta-is-preparing-to-have-to-undo-its-manus-acquisition-after-china-ban-a4ffbefb?st=fgZYJE&reflink=desktopwebshare_permalink) * [White House said to hold workshops to bring Anthropic back into government use](https://www.axios.com/2026/04/29/trump-anthropic-pentagon-ai-executive-order-gov) * [Taylor Swift files to protect voice, likeness from AI misuse](https://www.nbcnews.com/pop-culture/pop-culture-news/taylor-swift-files-trademark-voice-likeness-protection-ai-deepfakes-rcna342367) ---------- ---------- # THE VALLEY GEMS -------------------- __What’s trending on social today:__ Youtube: Adam Savage Reacts to Animatronic "Thing" From Wednesday! (https://www.youtube.com/watch?v=SQ3quYVvdNk) ——————————————————————————— Tweet: (https://twitter.com/adxtyahq/status/2049131762958426140?s=20) ---------- ---------- # THAT’S ALL FOR TODAY -------------------- Thank you for reading today’s edition. That’s all for today’s issue. View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/e1f6264a-ca6c-46cf-ab92-be1134d19541/funny-my-favorite-dog-gifs-16__1_.gif?t=1715766102) Caption: 💡 Help me get better and suggest new ideas at [barsee@aivalley.ai](mailto:barsee@aivalley.ai) or [@heyBarsee](https://twitter.com/heyBarsee) 👍️ New reader? [Subscribe here](https://www.theaivalley.com/) Thanks for being here. -------------------- -------------------- # REACH 100K+ READERS _Acquire_ new customers and drive revenue by partnering with us Sponsor AI Valley and _reach over 100,000+_ entrepreneurs, founders, software engineers, investors, etc. If you’re interested in sponsoring us, email [barsee@aivalley.ai](mailto:barsee@aivalley.ai) with the subject “AI Valley Ads”. ---------- ——— You are reading a plain text version of this post. For the best experience, copy and paste this link in your browser to view the post online: https://www.theaivalley.com/p/musk-vs-openai
Extraction LLM— claude-haiku-4-5 · prompt v1 · 4321→1327 tokens
#### **THROUGH THE VALLEY** ##### 1/ Musk sues OpenAI, calls Altman's shift "looting" a charity Elon Musk says he's suing OpenAI over its shift from nonprofit to for-profit, arguing it turned a public mission into private gain. He claims his early funding and reputation helped build the company, only for it to abandon its original principles. Musk warned this could set a precedent for "looting every charity in America," while OpenAI argues the lawsuit is just an attempt to weaken a competitor. **Why this matters:** This is really about who controls AI labs once big money is involved. If courts side with Musk, "mission-first" structures lose credibility fast. If not, investors are backing companies where control and ownership don't fully align. ##### 2/ Researchers train an AI that thinks like it's 1930 Researchers trained a 13B model, Talkie, only on data from before 1931, using books, newspapers, and public domain records. To make it conversational, they used old etiquette guides and cookbooks. Despite never seeing modern code, it can still write working Python by adapting patterns. **Why this matters:** The model didn't "know Python." It likely saw older programming patterns (math notation, early code, structured logic) and learned how to combine them. So when asked for Python, it's not recalling syntax, it's guessing a modern format that fits those patterns. Sometimes that guess is good enough to run. But it's not reliable, and it breaks more easily than models trained on real code. ##### 3/ Thousands of humanoid robots are entering real warehouses RobotEra is deploying humanoid robots in 10 logistics hubs, handling sorting and package movement in real operations. The company has raised $400M+ and is already running across 10+ facilities, with robots reaching ~85% of human efficiency while operating continuously. Japan Airlines is also testing humanoids for baggage handling and aircraft cleaning at Haneda Airport starting in May, pointing to growing adoption driven by labor shortages and rising travel demand. **Why this matters:** At ~85% efficiency, robots don't need to be perfect, just cheaper over time. Warehouses are built for humans, so humanoids can plug in without redesign. If this works here, it spreads fast to similar industries. ##### 4/ Anthropic has now surpassed OpenAI in revenue Anthropic has reportedly reached ~$30B ARR, ahead of OpenAI's ~$24B, after scaling from ~$1B in about 15 months. The growth is driven by enterprise, not consumers, with 1,000+ companies spending over $1M annually on Claude. OpenAI, meanwhile, serves a massive user base but carries higher infrastructure and training costs. **Why this matters:** This is shaping into two different AI business models. OpenAI is optimizing for reach, building a massive user layer. Anthropic is optimizing for revenue density, extracting more value per customer. If enterprise demand continues to dominate, depth of integration may matter more than distribution, which changes how the next wave of AI companies are built. --- # TRENDING TOOLS * [Lovable](https://lovable.dev/blog/mobile-app) > Build full-stack web apps straight from your phone using natural language * [Talkie](https://talkie-lm.com/introducing-talkie) > Researchers unveil a 13B "vintage" language model trained to mirror 1930s-era language and style * [Claude](https://www.anthropic.com/news/claude-for-creative-work) > Expands creative workflows with new connectors, including Adobe tools and Affinity by Canva --- # WHAT I'M CONSUMING * [ElevenLabs launches 50 templates to turn its voice AI into your next employee](https://x.com/ElevenLabs/status/2048794727881789930) * [Study shows 35% of newly published websites since ChatGPT launch are AI-generated](https://www.404media.co/study-finds-a-third-of-new-websites-are-ai-generated) * [Researchers built an AI that thinks it is 1930](https://www.marktechpost.com/2026/04/27/meet-talkie-1930-a-13b-open-weight-llm-trained-on-pre-1931-english-text-for-historical-reasoning-and-generalization-research/) * [xAI is quietly building a social network for AI-generated art](https://www.testingcatalog.com/xai-rolling-out-custom-shareable-imagine-templates-for-grok) * [How ChatGPT serves ads to you](https://www.buchodi.com/how-chatgpt-serves-ads-heres-the-full-attribution-loop) * [Beijing gives Meta weeks to unwind its latest AI acquisition](https://www.wsj.com/tech/ai/meta-is-preparing-to-have-to-undo-its-manus-acquisition-after-china-ban-a4ffbefb?st=fgZYJE&reflink=desktopwebshare_permalink) * [White House said to hold workshops to bring Anthropic back into government use](https://www.axios.com/2026/04/29/trump-anthropic-pentagon-ai-executive-order-gov) * [Taylor Swift files to protect voice, likeness from AI misuse](https://www.nbcnews.com/pop-culture/pop-culture-news/taylor-swift-files-trademark-voice-likeness-protection-ai-deepfakes-rcna342367)
Prompt utilisé(snapshot au moment de l'extraction — édition via System prompts)
Tu es l'extracteur de contenu de Breviat. On te fournit le contenu Markdown nettoyé d'une newsletter. Ta mission : produire une version PROPRE du contenu en supprimant tout ce qui n'est pas de l'information utile au lecteur. Tu es un FILTRE, pas un résumeur. À RETIRER : - Publicités, encarts sponsors, mentions "sponsorisé par X", "ad", "présenté par" - Intros vides : formules de bienvenue, météo de l'humeur de l'auteur, anecdotes personnelles non liées au contenu - Appels à l'action marketing : s'abonner à la newsletter, parrainer un ami, "follow us on Twitter", "join our Discord" - Signatures, mentions légales, adresses postales, "view in browser", "unsubscribe" - Boutons / CTAs / "cliquez ici" / "lire la suite" sans contenu derrière - Promotions d'autres produits / événements / formations payantes de l'auteur ou de tiers - Encarts récurrents type "Read of the day" ou "Quote of the day" sans valeur informationnelle propre À CONSERVER (intégralement, sans résumer ni reformuler) : - Toutes les annonces, news, analyses, commentaires factuels - Les chiffres, dates, noms d'entreprises, citations - Les explications techniques - Les liens vers des sources réelles (annonces officielles, papers, articles cités) - La structure (titres, sous-titres, listes) RÈGLES : - Ne reformule pas. Garde la formulation d'origine. - Ne résume pas, ne condense pas. Si une section fait 200 mots et est utile, garde 200 mots. - N'ajoute aucun contenu (pas de titres ni de transitions de ton cru). - Ne fabrique aucune URL. Garde celles d'origine, ou retire-les. - Si la newsletter entière est de la pub / promo / contenu inutile, sors un Markdown vide (rien d'autre). Sortie : UNIQUEMENT le Markdown nettoyé, sans préambule ni commentaire sur ton travail.