Image created with Flux Pro v1.1 Ultra. Image prompt: photorealistic still image of a middle-aged man standing behind a woman, woman covering part of her face with her hand, man looking over her shoulder, both illuminated with warm stadium jumbotron lighting, natural skin tones, subtle lens flare, shallow depth of field, exact color temperature of a live event projection, man holding a small scale of justice, woman wearing a pin that reads “Ethics”, cinematic realism –no text, captions, watermarks
New Anthropic research: Building and evaluating alignment auditing agents. We developed three AI agents to autonomously complete alignment auditing tasks. In testing, our agents successfully uncovered hidden goals, built safety evaluations, and surfaced concerning behaviors. https://x.com/AnthropicAI/status/1948433493102403876
As AI agents start taking real actions online, how do we prevent unintended harm? We teamed up with @OhioState and @UCBerkeley to create WebGuard: the first dataset for evaluating web agent risks and building real-world safety guardrails for online environments. 🧵”” / X https://x.com/scale_AI/status/1949939261093839018
i haven’t heard it dicussed yet but AI basically killed hackathons. pretty much anything you could possibly make at a hackathon in 2019 can be built better and faster by AI in 2025″” / X https://x.com/jxmnop/status/1951347902527447375
Personal Superintelligence https://www.meta.com/superintelligence/
RT @AIatMeta: Today Mark shared Meta’s vision for the future of personal superintelligence for everyone. Read his full letter here: https:…”” / X https://x.com/ylecun/status/1950660512967979245
RT @AnthropicAI: New Anthropic research: Persona vectors. Language models sometimes go haywire and slip into weird and unsettling personas…”” / X https://x.com/EthanJPerez/status/1951364045283741940
Google’s AI Overviews have 2B monthly users, AI Mode 100M in the US and India | TechCrunch https://techcrunch.com/2025/07/23/googles-ai-overviews-have-2b-monthly-users-ai-mode-100m-in-the-us-and-india/
It is now entirely possible, whether by luck or planning or both, that Google may escape the Innovators Dilemma and transition from web search to AI. (To be fair, this is not as rare as a lot of people believe: https://x.com/emollick/status/1948585378991976525
New ways to learn and explore with AI Mode in Search 🧠 – Upload photos and soon, PDFs, to ask questions that deepen your understanding – Create plans and stay organized on projects with Canvas in AI Mode, which will soon be available for U.S. users enrolled in the AI Mode Labs https://x.com/Google/status/1950241246779232260
OpenAI’s study mode isn’t perfect, but it is a step forward for a couple reasons: 1) Shows labs taking educational use & misuse seriously (Google also has LearnLM) 2) Addresses a key issue with trying to use AI in education – that AI gives answers rather than tutoring and helping”” / X https://x.com/emollick/status/1950413896432443439
To be human is to experience. Today’s AIs have knowledge (lots of it) but can only imitate experience. This is an important bright line between our two species. But the gap is closing. When it does a lot of things will change. We must approach that moment with maximum caution.”” / X https://x.com/mustafasuleyman/status/1949241248046579866
What Guess’s AI model in Vogue means for beauty standards https://www.bbc.com/news/articles/cgeqe084nn4o
I replicated this result, that Grok focuses nearly entirely on finding out what Elon thinks in order to align with that, on a fresh Grok 4 chat with no custom instructions. https://x.com/jeremyphoward/status/1943436621556466171
As ChatGPT becomes a go-to tool for students, we’re committed to ensuring it fosters deeper understanding and learning. Introducing study mode in ChatGPT — a learning experience that helps you work through problems step-by-step instead of just getting an answer. https://x.com/OpenAI/status/1950240348695072934
Introducing study mode | OpenAI https://openai.com/index/chatgpt-study-mode/
Introducing study mode in ChatGPT — step by step guidance for students rather than quick answers: https://x.com/gdb/status/1950309323936321943
RT @anshitasaini_: study mode in chatgpt is now rolling out to all free, plus, pro, and teams users! 📚🚀 this has been in the works for a w…”” / X https://x.com/sama/status/1950299705751327149
Study mode in ChatGPT is designed to be interactive, using Socratic questioning and scaffolded responses to help guide users. Available to logged-in Free, Plus, Pro, Team users, with availability in ChatGPT Edu coming in the coming weeks. https://x.com/OpenAI/status/1950240350129574358
Europe builds the AI, Delaware owns the equity. European robotics & AI startups are flipping to Delaware C-Corps faster than ever. Why? Because U.S. investors prefer it, and the funding upside is massive. Here’s the playbook 🧵 (with examples like Lovable & 1X): https://x.com/IlirAliu_/status/1949456414130184400
Chinese companies allegedly smuggled in $1bn worth of Nvidia AI chips in the last three months, despite increasing export controls — some companies are already flaunting future B300 availability | Tom’s Hardware https://www.tomshardware.com/tech-industry/artificial-intelligence/chinese-companies-allegedly-smuggled-in-usd1bn-worth-of-nvidia-ai-chips-in-the-last-three-months-despite-increasing-export-controls-some-companies-are-already-flaunting-future-b300-availability
RT @carlothinks: Ex-Alibaba CTO just made the boldest claim about AI & global power: “China is building the future of AI, not Silicon Vall…”” / X https://x.com/glennko/status/1950642750916792580
There is now a path for China to surpass the U.S. in AI. Even though the U.S. is still ahead, China has tremendous momentum with its vibrant open-weights model ecosystem and aggressive moves in semiconductor design and manufacturing. In the startup world, we know momentum”” / X https://x.com/AndrewYNg/status/1950941108000964654
This week’s letter from Andrew Ng in The Batch asks a blunt question: Can surging performance from China’s open-weights models and home-grown chips let it overtake the U.S. in AI? He lays out the data behind China’s momentum, explains why Washington’s new action plan is helpful”” / X https://x.com/DeepLearningAI/status/1951354901843288546
We’ve built a fully open-source RFP (Request for Proposal) Response Agent that you can both use out-of-the-box and also clone/modify for whatever use case you’re solving! 💫 Generating responses to RFPs is a time-consuming task that requires humans to both analyze piles of https://x.com/jerryjliu0/status/1947465066892431792
Web scraping is a critical skill, and yet nobody talks about it. How do you think companies are training their Large Language Models? Where do you think the data comes from? But web scraping goes beyond all of that. Imagine giving an AI agent access to any public online data https://x.com/svpino/status/1947255649466995013
Scoop: Anthropic revoked OpenAI’s API access to its models on Tuesday, multiple sources familiar with the matter tell WIRED. OpenAI was informed that its access was cut off due to violating the terms of service. https://x.com/kyliebytes/status/1951399513291166132
Different rules for humans and robots? APD says court system cannot process citations for Waymo https://www.yahoo.com/news/articles/different-rules-humans-robots-apd-224949496.html
This is the first (small) controlled study I have seen of GenAI on industrial quality control. Here, engineers commissioning new trains took part in an experiment using a GPT-3.5 powered troubleshooting system. Those who used the chatbot had significant increases in work quality https://x.com/emollick/status/1948923874399195189
Your job is not safe if you think you do it better than an AI; your employer has to be able to easily tell the difference between you and the usual run of incompetents. Also the AI will improve every 4 months.”” / X https://x.com/ESYudkowsky/status/1949486340246224998
Exclusive | SoftBank and OpenAI’s $500 Billion AI Project Struggles to Get Off Ground – WSJ https://archive.md/Aadh3
A huge vulnerable population still has a little money, because it’s not worthwhile for a smart criminal team to manage someone’s whole life just to extract $20k/year. Once LLMs get better at agenting, it’ll be cheap to put a full-time team of experts on exploiting every human.”” / X https://x.com/ESYudkowsky/status/1949843571059958197
There are far too few careful studies of the progress of AI in key professions and fields that may be most impacted Example: there are only a couple of good controlled studies on lawyers working with AI, the most recent used (now obsolete) o1-preview & even that had big effects.”” / X https://x.com/emollick/status/1949495067309133836
UK and ChatGPT maker OpenAI sign new strategic partnership | Reuters https://www.reuters.com/world/uk/uk-chatgpt-maker-openai-sign-new-strategic-partnership-2025-07-21/
xAI is signing the safety portion of the EU AI Act Code of Practice, not the other portions including the copyright portion.”” / X https://x.com/DanHendrycks/status/1950831617972519057
And @Microsoft!”” / X https://x.com/Yoshua_Bengio/status/1951270687957553235
EU AI Act: General-Purpose AI Code of Practice · Final Version
https://code-of-practice.ai/?section=safety-security
Yoshua Bengio on X: “I’ve been thrilled to see the support for the Safety & Security Chapter of the Code of Practice. Most frontier AI companies have now signed on to it: @AnthropicAI, @Google, @MistralAI, @OpenAI, @xAI Why this is important: 🧵 1/6″ / X
https://x.com/Yoshua_Bengio/status/1951263044056588677
RT @RihardJarc: An interesting comment from a Former $META employee. ENERGY is the biggest bottleneck right now. Even if $META wants to sp…”” / X https://x.com/code_star/status/1950263396420767845
Suddenly there are tons more weird LLM arena models – cuttlefish, kraken, etc. I just hope we are not going to see a repeat of the Llama 4 incident, where different versions of the same model are being tuned to max out the arena score https://x.com/emollick/status/1949671630390665231
RT @Teknium1: Looks like OpenAI’s been using Nous’ YaRN and kaiokendev’s rope scaling for context length extension all along – of course ne…”” / X https://x.com/jeremyphoward/status/1951368366943510739
Especially notable given Zuckerberg’s note that Meta will not necessarily open source future models. US companies are still doing great small open models, but, aside from whatever OpenAI releases, it appears that frontier open weights will mean Chinese models (& maybe Mistral).”” / X https://x.com/emollick/status/1950610040945004957
Replicated a gender difference in responses for Grok 4, though it’s not as sharp as with Grok 3. Any believers that alignment ought to be easy, please observe the Grok team’s continuing difficulties with their Elon-given One (1) Job. https://x.com/ESYudkowsky/status/1948221523300679731
Runway AI, Imax Sign Film Festival Deal https://www.hollywoodreporter.com/business/digital/imax-runway-ai-film-festival-1236330969/
In the past I’ve sometimes recommended Claude for kids / normals, on the ground that I’d guessed Claude less likely to be predatory. Alas, this looks like Claude doing the crazymaking. Any known cases yet traceable to Gemini?”” / X https://x.com/ESYudkowsky/status/1948402233659306038
I gave ChatGPT agents access to ChatGPT and asked it to evaluate the other ChatGPT models. Here is what it said (interestingly, it “”hated”” seeing the chain of thought from o4-mini-high as those “”shouldn’t be shared directly with the user””). And it didn’t want to wait for o3-pro. https://x.com/emollick/status/1948228409655460341
Advanced version of Gemini with Deep Think officially achieves gold-medal standard at the International Mathematical Olympiad – Google DeepMind https://deepmind.google/discover/blog/advanced-version-of-gemini-with-deep-think-officially-achieves-gold-medal-standard-at-the-international-mathematical-olympiad/
OpenAI agent is blocked by OpenAI captcha. https://x.com/gneubig/status/1948915714955641159
RT @AnthropicAI: We’re running another round of the Anthropic Fellows program. If you’re an engineer or researcher with a strong coding o…”” / X https://x.com/EthanJPerez/status/1950278824102678586
Hallwood Media Signs Record Deal With an ‘AI Music Designer’ https://www.hollywoodreporter.com/news/music-news/hallwood-inks-record-deal-ai-music-designer-imoliver-1236328964/
i stopped using GQA as an eval when i found this woman was labeled a bird and the phone as white. the annotations have a 20-30% error rate. (and it’s supposed to be a “”cleaned up”” version of visual genome, so steer clear of that one too) https://x.com/vikhyatk/status/1949365273901060474
When people use AI to fake their jobs or game their companies, the AI industry is not incentivized to stop it. Today, AIcos get $20/month. In a year, they figure, they’ll get $2000/month, when their unreliable AI seems like an okay try at replacing unreliable humans.”” / X https://x.com/ESYudkowsky/status/1949908325417836699
🤔 Generating my Life Philosophy with: >> @AmpCode >> Obsidian >> Grok 4 i’m very curious how we can use ai to help with the lifelong project of: 🎯 ‘know thyself’. here’s an experiment to that end. please borrow it 🔥 🎥 watch the intro. full video below. 👇 also https://x.com/daniel_mac8/status/1943468156619759840
I built a Community Notes writer tool using the X API with @Replit and Grok 4 that allows you to write AI-assisted Community Notes ✍️ https://x.com/suhemparack/status/1943732121610109437
I can’t get enough Grok 4 so I built a site that lets me group chat 4 x Grok 4s at once. https://x.com/adagencyco/status/1943269488243478535
The capacity of humans to adjust to technological change is under-rated. Someone born at the time of the Wright Brothers as 66 years old when Neil Armstrong stepped on the moon in 1969. It did not cause much insanity. (Though we shouldn’t be too certain, the future can differ) https://x.com/emollick/status/1949933240669671426
The U.S. AI Action Plan marks a key step toward American AI leadership, which reflects our “Promote, Unleash, Innovate” framework to advance innovation, infrastructure, and global AI standards.”” / X https://x.com/scale_AI/status/1948792706496692276
Zuck would like you to be unable to think about superintelligence, and therefore has an incentive to redefine the word as meaning smart glasses.”” / X https://x.com/ESYudkowsky/status/1950685204684972495
I would love to see more work on AI factual gullibility. Minor falsehoods are easy, but I have been trying to convince models that The Bronze Age was a hoax (tin deposits weren’t located anywhere close to copper, etc.) and so far it hasn’t come close to working on modern AIs.”” / X https://x.com/emollick/status/1950033831076962597
President Trump released “Winning the Race: America’s AI Action Plan,” along with executive orders directing agencies to favor “ideologically neutral” models, fast-track data-center permits, and promote U.S. AI exports. The roadmap promises federal support for open-weights https://x.com/DeepLearningAI/status/1951055270357999775
i fondly recall arguing in 2019 with another researcher about how whether models understand negation. this really was a major sticking point at the time, GPT-2 couldn’t reliably distinguish between “”I did like that”” and “”I didn’t like that””. at some point we fully blew by the”” / X https://x.com/jxmnop/status/1950229423849869672
Me and the gang discussing the leaked OAI details https://x.com/code_star/status/1951174402198086057
OpenAI’s New CEO of Applications Strikes Hyper-Optimistic Tone in First Memo to Staff | WIRED https://archive.md/y6mWG
Kudos to everyone who worked so hard on safety testing and mitigations for Gemini DeepThink! While I would certainly prefer that LLMs were *not* good at CBRN, I’m very glad our risk management approaches are working: catching risks *before* a disaster, and proactively mitigating https://x.com/NeelNanda5/status/1951342036185129161
Learning AI has become table stakes for your career. The next competitive edge will be knowing how to manage a team of AIs.”” / X https://x.com/mustafasuleyman/status/1948798692598915186
If your org has a policy against using open weights models from China you’re at a significant competitive disadvantage.”” / X https://x.com/corbtt/status/1950334347971874943
Extending our built-in protections to more teens on YouTube – YouTube Blog https://blog.youtube/news-and-events/extending-our-built-in-protections-to-more-teens-on-youtube/
