this is what a federal government completely controlled by Republicans looks like https://t.co/4ylCR9GBox

NOW: Massive delays at JFK Airport in NYC, crowds report waiting for hours in TSA lines Video by Diego Luzuriaga | Licensing @FreedomNTV Desk@freedomnews.tv https://t.co/n4mAO6tp7b

chat is this good https://t.co/9TIZXK48mN

chat is this good https://t.co/9TIZXK48mN

https://t.co/X9VNDgCL8W

https://t.co/X9VNDgCL8W

High quality STEP dataset from Shanghai QI ZHI institute. 23.9 TB in size https://t.co/Hlt3eF3ujm https://t.co/PNgUVgynSp

High quality STEP dataset from Shanghai QI ZHI institute. 23.9 TB in size https://t.co/Hlt3eF3ujm https://t.co/PNgUVgynSp

ELON MUSK: “To reach a petawatt, we’d build an electromagnetic mass driver on the Moon with robots like Optimus and humans. With no atmosphere and 1/6 gravity, payloads can reach escape velocity without rockets, drastically cutting costs. That could scale compute 1000× beyond a terawatt—I hope we live to see it.”

Elon Musk: "This chart explains why we need to build the TERAFAB." https://t.co/GF6q2UDneY

This would roughly be the scale of the Tesla/SpaceX TERAFAB compared to Tesla’s Giga Texas factory, which is already one of the largest buildings in the world. We are talking about an unprecedented amount of scale. https://t.co/SHL34TLNUd

⚡️This is Musk trying to build the industrial heart of a post-human power system. He is no longer talking like a car CEO or even an AI founder. He is talking like someone trying to compress the entire distance between thought, silicon, energy, and empire. TERAFAB is the visible artifact. The real objective is a self-reinforcing loop where chip design, mask creation, fabrication, testing, and redesign all live inside one accelerating machine. The next real bottleneck in AI is iteration speed at the substrate level. Whoever shortens the path from idea to chip to improved chip starts compounding faster than everyone else. At that point, compute stops being a purchased input and becomes an evolving sovereign capability. That is why this goes way beyond “new fab.” This is a bid to stop renting the future from outside foundries and start owning the recursion engine itself. Reports on TERAFAB explicitly tie it to Tesla, SpaceX, and xAI together, which tells you the real architecture is vertical integration across energy, hardware, AI systems, robotics, and eventually space infrastructure. The space language is the tell. When Musk says the goal is a trillion watts of compute per year and that much of it has to go to space because U.S. electricity is only about 0.5 TW, he is declaring that Earth is already too small for the civilization he wants to build. That is a species-scale doctrine. He is saying serious AI abundance eventually requires off-world energy and compute, and that the winners of this century will be the people who build the bridge first. The real view is simple. The arc is dead serious. The execution risk is enormous. Coverage notes there is no clear operating timeline yet and outside estimates already put the effort above $20 billion. But that does not change the signal. If this works even halfway, it becomes one of the most strategically important industrial projects in the world. If it fails, it still reveals the real direction of travel. The future belongs to whoever controls energy, compute, and the speed at which compute can redesign itself.

So, America… here you are. From a model of democracy to committing war crimes. I am amazed at how little time it took to get here, just one year. I am not surprised by Trump’s ultimatum to strike Iran’s power plants if the regime does not open the Strait of Hormuz. After all, it is no coincidence that his idol is Vladimir Putin, not Winston Churchill. What amazes me is how U.S. politicians and military leaders are allowing a deranged man to turn the United States into russia. I am amazed at how so many people, raised on the values of the free world, have turned into thousands of Marco Rubios, wearing shoes that do not fit, yet too afraid to say so. I am amazed that only a few have the dignity to stand up, while everyone else sits quietly and is ready to forgive Trump for protecting pedophiles, for constant lies, for violations of the Constitution, and now even for war crimes. And they remain silent not because they face death or imprisonment for speaking out, but simply because they fear losing their positions. It turns out that for the American establishment, losing a position is more frightening than war crimes. How did you come to this, America? How? Attacking civilian infrastructure in Iran will not solve the problems of Trump’s military operation. The old fool is too arrogant to learn from the experience of his friend Vladimir in Ukraine, who has been bombing our power plants for four years and has achieved nothing but greater resistance and anger among civilians toward the occupiers. The great American absurdity currently holding the office of President of the United States always chooses the most foolish option available. Right now, Trump simply does not understand that Iranians are not Americans. After years under the Ayatollah regime, they are used to hardship and daily struggles, and one more will not break them. Trump already betrayed the Iranian people once when he called for an uprising and then failed to support them, allowing it to be drowned in blood. And now, instead of seeking support from the people against the regime, this foolish old man wants to strike civilians to secure access to oil. This is an extremely cynical and vile plan. A war crime for the sake of oil. This war began as a fight of good against absolute evil, but it will become a disgrace for Trump and the United States. America, are you okay with this? Source: Translated and adapted from Serhii Marchenko

New research proves that current AI agent groups cannot reliably coordinate or agree on simple decisions. Building teams of AI agents that can consistently agree on a final decision is surprisingly difficult for LLMs. But problem is that developers frequently assume that if you have enough AI agents working together, they will eventually figure out how to solve a problem by talking it through. This paper shows that this assumption is currently wrong. Even in a friendly environment where every agent is trying to help, the team often gets stuck or stops responding entirely. Because this happens more often as the group gets bigger, it means we cannot yet trust these agent systems to handle tasks where they must agree on a correct answer. ---- Paper Link – arxiv. org/abs/2603.01213 Paper Title: "Can AI Agents Agree?"

This is the very simple case to use Grokipedia over Wikipedia. Just information, no spin. https://t.co/ze5SpgSIId

Frank Herbert speaks of this https://t.co/CpEao09ZA2

Art Deco Gotham vibes https://t.co/HBvjGvYchQ

Apps inside ChatGPT are the way to get your business discovered by 900M+ people using it, but building them is still too hard. So we built Fractal: @Lovable for ChatGPT Apps It’s the fastest way for business owners, with or without a technical team, to ship exceptional ChatGPT apps

Which story will senators rather tell their grandkids in 20 years? (A) “We kept working without a break for weeks on end; it was hard, but we passed the SAVE America Act, and it saved America” OR (B) “We could’ve saved America, but instead we took a two-week recess” https://t.co/u5ja1Ipl4m

Very different philosophies for skills in Codex versus Claude Code OpenAI seems to conceive of skills functionally, mostly matter-of-fact technical references for Codex. Claude skills are more about giving the AI approaches to problems See the difference in skill creator skills https://t.co/Gq3hcf1In0

Sakana AIは、読売新聞社と共同でSNS空間での中国による対日批判を分析しました。独自に開発したAI技術を搭載する当社のシステムが、SNSの膨大なデータから文脈やニュアンスを深く読み取り、批判投稿とナラティブを抽出。その構造を可視化し、さらに実用的な仮説構築までを実行しました。 【記事】 読売新聞（１） https://t.co/PpIpqfFdOv 読売新聞（２） https://t.co/wGCJx0gtJ9 今回の分析に用いたSakana AIの独自AI技術には三つの特徴があります。第一に、投稿文の文脈、ニュアンスからナラティブを抽出することです。例えば「高市首相の誤った発言の撤回を要請」という投稿から「台湾問題への介入と内政干渉」というナラティブを抽出しましたが、これは「台湾」というキーワード検索では発見できないものです。 第二に、独自開発したノベルティー・サーチ技術です。3種類の異なる大規模言語モデル（LLM）が集合知的に推論を重ねてSNS上の重要情報を探索し、粒度の高いナラティブを抽出します。それをさらに抽象度の高いグループにまとめて階層的に可視化することで、情報空間の大きな流れを詳細に把握できます。 第三に、仮説構築です。抽出・分類されたナラティブから、AIが仮説を無数に生成します。判断過程や具体的なデータなどの根拠も示されるため、分析担当者はその内容を精査し、再度分析を指示するなどして、信頼性が高いと考えられる仮説を絞り込むことができます。 今回は、SNSの計110万件にのぼる膨大な投稿の分析から、複数の仮説が導き出されました。このうち、「高市首相の国会答弁後、中国が統一的な対日批判戦略を検討してから大規模な対日批判を開始した」という仮説について、読売新聞が日中双方の政府関係者らに取材し、専門家の意見も踏まえて検証・裏付けを行いました。 AIは人間には見出せないインサイトを発見し、人間はそれを見てAIと対話しながらさらなる分析や対策を検討する－－今回の読売新聞社との共同研究はそうした新しい分析の実践の形を示したと考えています。 防衛・インテリジェンス領域においては、本調査で対象とした認知戦をはじめとして、「情報力」の果たす役割がかつてないほど大きくなっています。 Sakana AIはこうした背景も踏まえ、「金融」と並ぶ注力領域として「防衛・インテリジェンス」分野を位置付け、最先端のAI技術を実装する取組を進めています。日本発のAI開発企業として、引き続き防衛・インテリジェンス領域でのAI実装を本格化していきます。

+1 more

We recently worked with The Yomiuri Shimbun to analyze more than a million social media posts to map out state-sponsored information campaigns. https://t.co/fBQQLg1glS Keyword searches are fragile for modern OSINT. To fix this, our team used an ensemble of different LLMs combined with our Novelty Search algorithm to extract underlying narratives purely from context. (e.g., The system successfully mapped posts demanding "a politician retract a statement" to the broader, hidden narrative of "Taiwan interference"). The system clusters these granular narratives hierarchically and generates testable hypotheses, citing specific evidence. Human journalists took the AI-generated hypotheses, interviewed real-world government sources, and verified the timeline of the coordinated campaign our system uncovered. Fascinating look at human-AI collaboration for intelligence analysis.

1/n 🦞 Introducing EurekaClaw💡 — a local-first AI research agent that captures your Eureka moments before they vanish. From idea → proof → experiment → paper — fully automated. Local-first. Zero data leak. 🔒 Try it: https://t.co/ZJlEYvkX9h

Anthropic showed older (2022) LLMs will give you less accurate answers if you seem less educated to the AI, but this issue has, as far as I know, been addressed in more recent models. https://t.co/iXnlBpYXLA https://t.co/ByclHjwywR

@XSquareRobot This looks incredible! Would love to have you guys demo this at AI Engineer Singapore in May, we have a robotics track and we'd love the robots running around :) https://t.co/LjY53iRZcn

cruising around 16K shapes at 120fps https://t.co/V2LPqrxaDL

Two MacBooks in my bottega https://t.co/MdFsgnBvXz

+2 more

@santa_kaus https://t.co/Pt34qekNRw