J.A.

@ImGinnyRobinson Looking for parties to attend?

Why is Elon Musk suddenly hanging out with demonic Howard Lutnick? 🤔

@MarioNawfal Fake

Tesla Optimus over here rolling perfect sushi like it's nothing. The competition? Robot waiters dancing and breaking dishes until security drags it off stage. One cooks, one dances.

@VraserX What happens with his eyes when he blinks 😂😂😂

I still don’t get the DLSS 5 hate. This looks insane. 🤯

@elonmusk There is no story here, just AI slop clips that look similar in style stitched together.

Minute-long story made w Grok Imagine

@elonmusk Why would anyone want to watch this? At least make something that's a bit funny or tells a story, otherwise it's just a waste of compute and energy.

Imagine

@SawyerMerritt @SpaceX Pretty sure this post won’t age well when Starship and the Heavy Booster fail to deliver on its government contracts.

I hate when people say @SpaceX is simply handed government contracts. No, SpaceX earns those contracts. They offer the lowest price, the best product and they execute. The Pentagon said last year that SpaceX has saved the government over $40 billion. One SLS launch costs billions, while one SpaceX launch costs ~$75M. SpaceX is an example of great American innovation, something that all Americans should cheer on.

@teslaownersSV And teleoperated…

Optimus 🤖 — useful, smart, and a little mischievous 😏

@KatieMiller You are almost incorrect 100% of the time, so I guess ChatGPT beats you then.

ChatGPT is effectively correct only about 60% of the time. When your train your model on Internet garbage, this is the predictable result.

@VraserX Because he knows it will never happen.

This isn’t a flex, it’s a warning. “Exceed everyone combined” in AI means monopoly-level power over one of the most dangerous technologies ever built. You’ve spent years criticizing OpenAI for this exact risk… and now you’re openly aiming for it yourself. How is that better?

@Whale_Guru @thunderf00t Only Trump s stupid enough to beg Switzerland for naval support.

TRUMP BEGGED HIS ALLIES FOR WARSHIPS TODAY. EVERY SINGLE ONE REFUSED. This is not just embarrassing. This is the most dangerous thing that has happened to American power in decades. Here is what actually happened. Trump spent 12 months calling NATO "obsolete." He threatened Canada with annexation. He slapped tariffs on every European ally. He abandoned Ukraine. He called Germany weak and France irrelevant. He insulted world leaders to their faces publicly. Then Iran mined the Strait of Hormuz. 20% of global oil. Shut down. Overnight. Trump calls his allies. Needs warships. Needs backup. France: no. Germany: no. Norway: no. Canada: never. Japan: officially no. Australia: we'll think about it. UK was actually preparing a carrier — Trump told them "don't bother" last week. Now he's begging for warships and nobody is picking up. The real danger is not Iran. The real danger is America learning it cannot fight a global war alone. $21 billion spent. 14 Americans dead. Oil at $102. Baghdad evacuated. Iran still attacking. Week 3. No exit. No coalition. No plan. You spent a year breaking your alliances. Now you need them. They listened to you.

@teslaownersSV After ten years of development, it hasn’t even achieved full Earth orbit.

Starship is the first rocket ever designed to go to the Moon and Mars. It’ll have 42 engines.

@elonmusk With Elon it’s always next year, soon, probably, likely….. The worst planner on earth 😂

True. New Roadster unveil probably in late April.

@abxxai Lol, people fabricate answers too

BREAKING: 🚨 Someone just tested 35 AI models across 172 billion tokens of real document questions. The hallucination numbers should end the "just give it the documents" argument forever. Here is what the data actually showed. The best model in the entire study, under perfect conditions, fabricated answers 1.19% of the time. That sounds small until you realize that is the ceiling. The absolute best case. Under optimal settings that almost no real deployment uses. Typical top models sit at 5 to 7% fabrication on document Q&A. Not on questions from memory. Not on abstract reasoning. On questions where the answer is sitting right there in the document in front of it. The median across all 35 models tested was around 25%. One in four answers fabricated, even with the source material provided. Then they tested what happens when you extend the context window. Every company selling 128K and 200K context as the hallucination solution needs to read this part carefully. At 200K context length, every single model in the study exceeded 10% hallucination. The rate nearly tripled compared to optimal shorter contexts. The longer the window people want, the worse the fabrication gets. The exact feature being sold as the fix is making the problem significantly worse. There is one more finding that does not get talked about enough. Grounding skill and anti-fabrication skill are completely separate capabilities in these models. A model that is excellent at finding relevant information in a document is not necessarily good at avoiding making things up. They are measuring two different things that do not reliably correlate. You cannot assume a model that retrieves well also fabricates less. 172 billion tokens. 35 models. The conclusion is the same across all of them. Handing an LLM the actual document does not solve hallucination. It just changes the shape of it.

@MELANIATRUMP Agree, one day is more than enough!

MELANIA The Record-Breaking Film Only on Prime March 9th

@sama This doesn’t look good, Sam! I had higher thoughts about you. 😢

Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome. AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement. We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only. We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements. We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place.

@MarioNawfal John Koenig will kick Elons ass !!

𝕏’s superhero off to Moonbase Alpha.

@KatieMiller @AOC Do you know what a map is?

@AOC Are you sure you know where it is on a map?

No war with Iran.

@VraserX Moonbase Alpha with me in it.

If your favorite show gets cancelled, soon you will just generate Season 2 yourself. And yes, it will be better than the official one. What show do you resurrect first

@daniel_stahre @Pro_Tradings Ngn gammal övervintrad EAPare?

@Pro_Tradings Vem är den där idioten?

🤙

@MarioNawfal Nothing wrong with Poland, but then it's weird that there is no mass immigration inside the EU to Poland. How do you explain that?

🇵🇱"Zero refugees, zero terrorist attacks. Lowest unemployment in the EU. Highest GDP after COVID. Call us racist, I don't care. I care about my family and my country." Be like Poland indeed. x.com/theblaze/statu…