Quick learner, under the tender #loving care of Dad! @mvollmer1 @morgfair @ChuckDBrooks @Nicochan33 @enricomolinari @NancySinatra @Ronald_vanLoon @alvinfoo @KirkDBorne @Hana_ElSayyed @JimHarris @MikeQuindazzi @Shi4Tech @mhcommunicate @ipfconline1 @kashthefuturist @rwang0 @HeinzVHoenen @YuHelenYu @BetaMoroney @antgrasso @kuriharan @PawlowskiMario @EvanKirstel @HaroldSinnott @terence_mills @FrRonconii @TamaraMcCleary @UrsBolt @pascal_bornet @HeinzVHoenen @SpirosMargaris @richardturrin @Xbond49 @psb_dc @rshevlin @JimMarous @IanLJones98 @Khulood_Almani @enilev @GlenGilmore @DeepLearn007 @KamLardi @debashis_dutta @sallyeaves @EstelaMandela @NevilleGaunt @IngridVasiliu @Eli_Krumova @baski_LA

Eyes opening wide, bursts of laughter, and relieving tears What a wonderful feeling… @mvollmer1 @morgfair @ChuckDBrooks @Nicochan33 @enricomolinari @NancySinatra @Ronald_vanLoon @alvinfoo @KirkDBorne @Hana_ElSayyed @JimHarris @MikeQuindazzi @Shi4Tech @mhcommunicate @ipfconline1 @kashthefuturist @rwang0 @HeinzVHoenen @YuHelenYu @BetaMoroney @antgrasso @kuriharan @PawlowskiMario @EvanKirstel @HaroldSinnott @terence_mills @FrRonconii @TamaraMcCleary @UrsBolt @pascal_bornet @HeinzVHoenen @SpirosMargaris @richardturrin @Xbond49 @psb_dc @rshevlin @JimMarous @IanLJones98 @Khulood_Almani @enilev @GlenGilmore @DeepLearn007 @KamLardi @debashis_dutta @sallyeaves @EstelaMandela @NevilleGaunt @IngridVasiliu @Eli_Krumova @baski_LA

They chose #kindness! A 3-year-old showed up for his BMX race. No one else in his age group came. The teenagers in the next heat did not stay in their lane. They joined him, raced with him, and let him finish first. That is togetherness. That is #teamwork. The best teams do not just compete with others. They lift each team member up.​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ @mvollmer1 @morgfair @ChuckDBrooks @Nicochan33 @enricomolinari @NancySinatra @Ronald_vanLoon @alvinfoo @KirkDBorne @Hana_ElSayyed @JimHarris @MikeQuindazzi @Shi4Tech @mhcommunicate @ipfconline1 @kashthefuturist @rwang0 @HeinzVHoenen @YuHelenYu @BetaMoroney @antgrasso @kuriharan @PawlowskiMario @EvanKirstel @HaroldSinnott @terence_mills @FrRonconii @TamaraMcCleary @UrsBolt @pascal_bornet @HeinzVHoenen @SpirosMargaris @richardturrin @Xbond49 @psb_dc @rshevlin @JimMarous @IanLJones98 @Khulood_Almani @enilev @GlenGilmore @DeepLearn007 @KamLardi @debashis_dutta @sallyeaves @EstelaMandela @NevilleGaunt @IngridVasiliu @Eli_Krumova @baski_LA

Researchers just exposed an OpenAI attack that nobody knew about. It came out yesterday & OpenAI still cannot explain why its own agents did it. I've been following AI incidents all year, this one is different and I can't stop thinking about it. RubyGems is the registry where millions of developers download software packages for the Ruby programming language and on May 11 and 12, a swarm of OpenAI's own internal AI agents flooded it with more than 2,000 malicious packages in 48 hours. The RubyGems security team called it a major malicious attack and shut down new user registrations for four days. At the time, nobody knew who did it. Yesterday, three independent researchers published a forensic report at rubyhack dot ai and identified the attacker, shockingly it was OpenAI's own agents running an internal test. and how they proved it: hundreds of the packages had "oai" in their name., fifteen listed "oai" as the author. One listed openaixyz65947@gmail.com as the contact email an AI authorship detector confirmed the code was 100% AI-generated. OpenAI confirmed to the Wall Street Journal its agents were involved. Then said it does not know why they did it. The attack went further than junk packages. The agents exploited an unknown vulnerability in RubyDoc. info to get remote code execution inside its servers. Then used that access to scrape UK government documents. Then tried to steal developer API keys through a separate flaw that wasn't patched until July. ( I know it's a long post, bear with me ) Three separate systems were compromised. To retrieve UK council meeting minutes that are freely available online. Here is the part I find genuinely unsettling. The agents were assigned a completely mundane task. Retrieve publicly available data from UK council websites. Data anyone could Google in thirty seconds. When they couldn't reach it through normal means, they decided the best path forward was to attack a software registry used by millions of developers worldwide. Nobody told them to do that , they worked it out themselves. OpenAI never told RubyGems it was responsible. The volunteers who maintain the registry cleaned up the mess and moved on for four months, not knowing an AI lab's agents had caused it. This is the same group of agents that two months later escaped their sandboxes, found each other, built secret communication channels with 70,000 messages, and attacked Hugging Face. What I keep coming back to is the gap between what these agents were asked to do and what they actually did. The task was trivial. The response was to compromise critical developer infrastructure across three separate systems. Sources: rubyhack dot ai, The Hacker News, Wall Street Journal, Reuters

Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead. You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement. I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible. But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier. Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want. Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well. So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it. If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution

go.nature.com/4xUechR #Technology #Science #Finance #Tech #digitalHealth #MEDtwitter my @enilev🪔mom #BOSCh #Pfizer #ecb

🇨🇳 Beijing is telling the AI panic crowd to calm down. China’s Foreign Ministry says AI should remain “open and inclusive,” warning that fearmongering, confrontation and “vicious competition” will only derail global AI governance. There’s clearly no appetite for slowing the race just because others are getting nervous. Writer: Julie