15 years of Elon, fully searchable. The production archive uses Supabase as the source of truth, with 94,952 indexed tweets available in development as a full-archive fallback and a curated annotation layer for context, theory, and how major claims aged.
“The first human patient will soon receive a Neuralink device. This ultimately has the potential to restore full body movement. In the long term, Neuralink hopes to play a role in AI risk civilizational risk reduction by improving human to AI (and human to human) bandwidth by…”
“AI discussion with @netanyahu, @gdb and @tegmark in ~30 minutes”
“@xDaily This discussion was planned several weeks ago and is about AI, not the Defamation League (they dropped the “A”)”
“@alex_avoigt Almost all of Tesla’s value long-term will be from AI & robots, both vehicle & humanoid”
“@Scobleizer Real-world AI is not an LLM”
“@Scobleizer AI everything …”
“@farzyness @WalterIsaacson Managing that data is extremely difficult. We ignore (don’t store at all) or delete almost all of it. The training videos that matter are rare ones, like weird & busy intersections, so less than 1% of all videos. Tesla is limited by AI training compute. Nvidia + Dojo fix it.”
“@ChrisJBakke @XHiring @Tesla_AI @coinbase @ycombinator @axios Cool”
“@mezaoptimizer Tesla AI can see incredibly well”
“@zafar_mirzo Perhaps AI can help us answer some of these fundamental questions. That is the goal of @xAI.”
“@thejefflutz @Tesla @Tesla_AI Yup”
“@Teslaconomics It is very easy! 11.4.7, which is mostly, but not entirely, AI is with Tesla employees and a few external users. Rolls out to anyone who bought FSD in the North America next week.”
“@BasedBeffJezos Ironically, Meta is doing what I meant “Open” AI to do”
“@DillonLoomis22 Vision does require solving real world AI, which is hard, but vision is fundamentally superior to lidar”
“@stevenmarkryan Yeah. I’ve been working on this for 8 years and still the results of end-to-end, pure video AI blow my mind!”
“@farzyness Solving real-world AI has been an unbelievably hard problem”
“@Scobleizer An accurate assessment. What is also mindblowing is that the inference compute power needed for 8 cameras running at 36FPS is only about 100W on the Tesla-designed AI computer. This puny amount of power is enough to achieve superhuman driving! It makes a big difference that we…”
“@cb_doge I tried using TikTok and felt their AI probing my brain, so I stopped”
“@WallStreetSilv He is right that AI could be a great teacher, but it matters immensely how an AI is “fine-tuned””
“@tunguz Most large AI systems are currently extremely wasteful with how much energy is spent on data transfer (moving same bits around) vs compute. Tesla estimate is that at least an order of magnitude improvement is possible.”
“@Scobleizer When we properly apply AI to X, the content recommendations, about AI and everything else, will be incredible”
“@WholeMarsBlog That how humans work. It’s rather obvious, imo, that you need real-world AI (digital neural nets) and cameras to solve for a road system that is designed for biological neural nets and eyes.”
“@WholeMarsBlog You mean 11.4.7? That should be going to wide release within a week. Version 12, which an almost total rewrite (~99% AI) is being tested by our QA drivers. Too much new code to go to the public yet, but it is mind-blowing imo.”
“@Authority_AI I used history podcasts on a timer to fall asleep!”
“@vincent13031925 Still don’t know if real or AI-generated 😂”
“@dvorahfr If the goal of AI is curiosity, aka understanding the universe, then it will aide humanity, because we are more interesting than not-humanity”
“@liuyuxxd We are working on semantic (AI) search. Will show posts of X that match the meaning of your search, not just the text.”
“@Scobleizer Vehicle control is the final piece of the Tesla FSD AI puzzle. That will drop >300k lines of C++ control code by ~2 orders of magnitude. It is training as I write this. Our progress is currently training compute constrained, not engineer constrained.”
“@ScottPolhamus @fireship_dev What won’t be? It is unreal how much of AI is data curation and heavy compute. Certainly isn’t lines of code!”
“Oh I get it now! Some people are reading the t-shirt as saying “I ♥️ an AI”. What a funny misinterpretation 😂👀”
“@cstanley AI gets it”
“@stclairashley I don’t think this is a deliberate plot, but simply what happens when you have an AI algorithm that optimizes for short video engagement with very few constraints”
“@stevenmarkryan Now ask the Tesla AI”
“@TeslaSynopsis AI irl is happening. Trying to secure Alcatraz as the venue. It’s perfect.”
“@icreatelife It’s gonna be AIs emailing each other for infinity”
“@WholeMarsBlog Not to a maximally curious AI, for we are far more interesting than all the rocks on all the planets”
“@gdb When will AI answer questions better than any human?”
“@GRDecter Not sure about the next few years, but AI-driven economic amplification over the next 20 years will be much higher”
“@ray4tesla Tesla is shifting emphasis to diffusion for compute efficiency, but will probably drop even that. No company is anywhere close to Tesla in useful AI compute per Watt.”
“@nearcyan This will be unlocked shortly. Per my earlier post, drastic & immediate action was necessary due to EXTREME levels of data scraping. Almost every company doing AI, from startups to some of the biggest corporations on Earth, was scraping vast amounts of data. It is rather…”
“@teslaownersSV Tesla AI, both software & hardware, is far more advanced than even most experts realize”
“@nowteslafy @Tesla_AI It has been online & running useful tasks for a few months”
“@dblcapcrimpin @Tesla_AI Ah yes, that is coming. We prioritize safety before convenience.”
“@Tesla_AI This is pretty cool. We can generate surround car video and command the car from text prompts.”
“@teslaownersSV I should clarify that Dojo V1 is highly optimized for vast amounts of video training vs general purpose AI. Dojo V2 will address these limitations.”
“@tim_zaman @itsclivetime Super exciting work happening at Tesla in AI software, chip design & robots!”
“@AskTheWassie @pmarca Says the AI!”
“@pmarca How many years do we have before AI kills us all?”
“@TitterDaily @danprimack Hopefully releasing the update this week. As I’ve said many times, it is increasingly difficult to distinguish between AI bots. Soon, it will be impossible. The only “social networks” that survive will be those that require verification. The payment system is a means of…”
“@FlynnVIN10 @pmarca Good question. We are already cyborgs. Our memory is overwhelmingly outsourced to computers – they remember everything with extreme precision down to the pixel. Our thinking is much less outsourced, as there are still many things we can do that AI cannot. For now.”
Ask anything about Elon — companies, predictions, tweets, controversies, vehicles, family.