Grok Proves It Can Improve Any Post
Hey Grok, can you improve this post? Yes, it can! (It improved this post).
AlphaFold 2 Marks 5 Years, Nobel, 3M Users
Thrilled to celebrate 5 years of AlphaFold 2! It’s now been used by over 3 million researchers around the world to accelerate their vital research - and it was an honour of a lifetime for our work to be recognised...
Gemma AI Uncovers Promising New Cancer Therapy Pathway
How a Gemma model helped discover a new potential cancer therapy pathway https://t.co/lMiONmkjfO @GoogleDeepMind @GoogleForHealth #ai #iot #5g #agenticai #ces2026 #Healthcare @_atanas_ @enilev @Fabriziobustama https://t.co/obWVS8SlMt
SketchVerify Enables Fast, Physically-Coherent Video Generation
"a major unlock for embodied AI, robot simulation, autonomous agents, and anyone who wants controllable video models that actually work in the physical world instead of just producing pretty hallucinations." That's what Grok says. It says more about why this work...
Tesla’s Autopilot Outperforms Human Drivers, Guests Amazed
Smart of Tesla to do. Lots of Tesla owners have never tried it. And now not only do they get to try it, but they get to take their Thanksgiving guests on drives to demonstrate it to others. It is...
AI Layoffs Risk Damaging Employer Reputation and Talent Pipeline
How AI layoffs could backfire on employers? https://t.co/dtqbOC75t4 #AI #artificialinteliigence #layoffs #LayoffsInIndia #jobloss
Tech Boom Threatens Life: The War on Humanity
If you want to spice up Thanksgiving with fresh anti-Big-Tech ideas, my brother's book "The War on Life" just launched in English: https://t.co/rc8zSHqAov Here's my take on his book. We're all aware that artificial intelligence and other tech is transforming...
Video-to-Video Diffusion Expands VTuber Self-Image Flexibility
reskinning myself playing electric guitar. canon rock for the vibes. the future of vtubing is bright -- video-to-video diffusion is about to make everyone's self-image a lot more flexible. https://t.co/OWBIA5HVFu
Stable Browsing Layer Solves Web Agents' Blocking and Scaling Issues
Here’s the real problem with web agents: → They get blocked CAPTCHAs, 429s, rate limits, fingerprinting, IP bans. Any repetitive pattern gets flagged fast. → They don’t scale Try running more than a handful of sessions and you burn IPs, crash browsers, or lose...
Whisper Thunder: World's Top Text-to-Video Model Revealed
There’s a new text-to-video model that’s #1 in the world. Who is Whisper Thunder? https://t.co/KAq9MwuMYt
AI Writing Evolves: Generate, Critique, Refine, Repeat
Learned via @kelvinckchan today The old : Input --> Generate The new: Input --> Generate --> Critique / Check --> Refine --> Output .... > Input https://t.co/izYA7LfcbH
Vercel Launches Open-Source AI-Powered Visual Workflow Builder
Yess! @Vercel just dropped their open-source visual agent and workflow builder! 🔥 ▪️ Built on https://t.co/P15EXKiAlw ▪️ Outputs “use workflow” code ▪️ AI text→workflow ▪️ AI SDK + Elements ▪️ Integrations: Resend, Linear, Slack Demo + free template in 🧵↓ https://t.co/Ef7raJpJ7K
Non‑invasive BCI Hats Poised to Rival Neuralink by 2027
Vinod has seen more brain/computer interface companies than anyone else. If he says non-invasive is better that is bullish. One of the companies he saw is making a hat that does pretty much what a @neuralink does today. “Think and compute,” is coming...
Inference Scaling Boosts LLM Accuracy From 15% to 52%
As we head into a long weekend, some of you may be looking for reading material. Good news is that Chapter 4 on inference scaling was just released earlier this week! This chapter introduces the core ideas behind inference scaling...
AI‑generated Artist Hits Billboard, Confirming Music Future
AI Music Prediction I got skewered for in 2019: AI will play a big role in entertainment and design — music and entertainment will be plentiful and personalized for you and your mood (TechCrunch; 2019). We’d have custom, AI-generated soundtracks...
Claude Opus 4.5 Gains Math Usability, Scores 21%
Claude Opus 4.5 is the first Claude I think is reasonably usable for decent math work (Claude interface is great for iterating minus the timeouts & mobile slowdowns) The big thing here though that I've noted using Opus 4.5 usage is...
20 Essential AI Tools to Master in 2025
🔥 20 must-try AI tools for 2025. This stack covers everything: 🤖 chat & reasoning 🎨 image gen 🎬 video gen 🗣️ voice/music 🔎 AI search 🛠️ automation 🧠 AI agents My top 3 right now: ChatGPT, Perplexity, Midjourney. What are yours? 👇 #AI #AITools #GenerativeAI #Automation #AgenticAI
KV Cache Speeds up AI After First Prompt
Your first question to an AI model takes a moment… But the next ones appear almost instantly, there’s a simple reason behind it. The model keeps a small snapshot of the work it already did. This is called the "KV Cache". When an AI...
Combine BM25 and Vectors with Reciprocal Rank Fusion
In this video, we upgrade our simple BM25 search engine into a Hybrid Search system using BM25 + vector embeddings with reciprocal rank fusion. What you will learn: ✔️ What hybrid search means ✔️ How to add embedding fields ✔️ How to use HNSW...
Human Trust, Governance, Skills Block AI Adoption
💬 Excited to share that I was recently quoted in this recent article from The Digital Project Manager exploring the real challenges PMs face when bringing AI into their work. The big takeaway? AI isn’t hard because of the tech...
Build a Hybrid Search in 1.5 Hours
Starting in 1.5 hrs! Build your own Hybrid Search https://t.co/E8PlCJkvJL
Open IP Fuels Huawei’s Massive AI SuperPod Breakthrough
The 𝗛𝘂𝗮𝘄𝗲𝗶 𝟮𝟬𝟮𝟱 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻 & 𝗜𝗻𝘁𝗲𝗹𝗹𝗲𝗰𝘁𝘂𝗮𝗹 𝗣𝗿𝗼𝗽𝗲𝗿𝘁𝘆 𝗙𝗼𝗿𝘂𝗺 once again underlined how openness in IP and sustained R&D investment can unlock the next wave of digital innovation. Among the Top Ten Inventions presented this year is the 𝗦𝗰𝗮𝗹𝗲-𝗨𝗽 𝗨𝗹𝘁𝗿𝗮-𝗟𝗮𝗿𝗴𝗲-𝗦𝗰𝗮𝗹𝗲 𝗦𝘂𝗽𝗲𝗿𝗣𝗼𝗗...
Google AI Finally Answers, Then Resets Conversation
Google Ai gave an answer this time without just ghosting with a Delete-convo-add .. next question will start a fresh chat. https://t.co/JaiWURMZMx
AI Bans Cripple U.S., Give China Unstoppable Advantage
Here's the rub. Let's say all the states ban AI toltally except for one. That state would end up out producing all the others 10x-1 within a decade. And possibly 1,000x-to 1. And I bet that the math even supports a...
Scaling AI Boosts Confidence, Not Factual Accuracy
AI models are getting stronger, but their confidence still grows faster than their certainty. More parameters expand what the system can talk about, not what it can truly verify. The space of possible answers grows faster than the ground beneath...
Bigger AI Models Boost Confidence, Not Factual Certainty
AI models are getting stronger, but their confidence still grows faster than their certainty. More parameters expand what the system can talk about, not what it can truly verify. The space of possible answers grows faster than the ground beneath those...
Future Tech Will Amplify Addiction—Act Now
@PAstynome When World Models really show up in spatial computing devices (in about two years) it will get far worse. And when brain computer interfaces show up it will get far worse again. So, we need to figure out addiction issues now.
TPUs Outperform NVIDIA Chips in Large‑scale 8‑bit Training Stability
@zephyr_z9 TPUs are much more stable on 8 bit training (aqt etc) than NVIDIA chips at massive scale Previous gen was a bit sensitive on topology but looks like less of an issue for ironwood
Curated AI Lists Bypass X’s Follow Limit
Following everyone in AI does makes the For You feed much, much smarter. Unfortunately for you X doesn’t let you follow 31,000 people. I wish you could see mine. I put them all into lists so you can get some of the...
Comparing GPT‑5.1‑Codex to GPT‑5.1‑Codex‑Max
@pagilgukey @JohnThilen @dwarkesh_sp @ilyasut Yes. With Codex I meant GPT-5.1-Codex versus GPT-5.1-Codex-Max
Open Federal Datasets on Hugging Face to Power Collaborative AI
@ScienceUnderSec Inspiring work. Let's put all the federal datasets on HF to unleash open and collaborative AI for science!
AI-Driven Dental Offices Boost Satisfaction, Signal Future
The dentists' office of the future. Here today at @TeamCloudberry. Customer satisfaction numbers are way up. Welcome to the automated dentists' office. The AI does all the grunge tasks other than putting their fingers in people's mouths. Customer service. Tracking of everything....
Spare Compute Needed to Accelerate Idea Testing
In some way, scaling is holding back progress. And either way these mega-size clusters are going to be useful. Right now, most of the capacity is used to do a crazy large run + serving existing customers. It would be good...
FLUX.2 Launch Sparks Early Holiday Excitement
Today's big model release is FLUX.2 from @bfl_ml. It's like Christmas before Thanksgiving lately, isn't it?
Explore Cutting-Edge Visual Positioning and 3D Scanning Tech
check out @multiset_ai (visual positioning system & object tracking platform) and @XGRIDS2023 (3d scanning devices)
Real‑time 3D Mapping Delivers X‑Ray Precision Localization
3d visual positioning experiment -- look at the alignment between the 3d mesh and the live camera view. Truly feels magical -- like x-ray vision. Workflow: 1. Scanned a street in 15 mins w/ xgrids 2. Localized against that scan at night, in...
Google's New IDE Blocks Payments, Not a Paywall
used google's new IDE for 5 minutes, was impressed, hit what I thought was a paywall but it's not a paywall, you just can't give google money? I don't understand https://t.co/cKUTuUQNsD
Clarity Beats Speed: Narrative Accelerates Growth
Marketing is the new bottleneck. I tossed that out as a quick thought. The replies came fast. The private messages went deeper. People wanted to hear more and understand exactly what I meant, so I turned it into a thread....
AI Breakthroughs Now Arrive in Just Five Years
@w3whq @kenwarner GANs were 2015ish, Denoising Diffusion Probabilistic Models were 2020ish, aka 5 years later. Timeline expectations are crazy these days!
AGI's Arrival Uncertain; Automation Fuels Economic Boom
After listening to @ilyasut I am even more convinced we won’t know when AGI gets here. But liked how he predicted humans will change as AI gets better. Economic boom times arrive as more people get into automating their businesses and...
GPT-5 Expected to Be Smaller Than GPT-4.5
@JohnThilen @dwarkesh_sp @ilyasut In addition, and that’s the important point, I think GPT-5 is smaller than GPT-4.5.
GPT‑5.1 and Gemini 3 Variants Share Identical Core Model
@JohnThilen @dwarkesh_sp @ilyasut I am speculating that all GPT-5.1 models (instant, thinking, Pro) are the same model but with different inference scaling budgets. Same for GPT-5 Codex. And Gemini 3 Pro and Gemini 3 Deep Think are probably also the same...
Dev Roles Shift From Coding to Code Evaluation
How do devs feel about this job change? Example below in Codex. Pros: you can kick off tasks from anywhere (just talked to one dev who started multiple codex tasks while getting into a cab), you get multiple versions to pull...
Scaling Pre‑training Hits Diminishing Returns for Future Generations
@GiorgioMantova @dwarkesh_sp @ilyasut I’d say this is the jump from last gen to current gen, but I think the argument is that further improvements will fizzle out in the next gen if we keep scaling pre-training. Ie it won’t give...
Flux.1-dev Ranks #2, Eagerly Awaiting Flux.2-dev
Flux.1-dev has been the second most liked model on Hugging Face just after Deepseek R1 so super excited to see the release of Flux.2-dev by @bfl_ml today! Download the weights or try the model (thanks to @fal) on @huggingface: https://t.co/kdmVlvdLZh Read the...
Reachy Mini Becomes My New Podcast Assistant
Reachy mini is my new podcast assistant! Coming soon with @ti_morse... https://t.co/VfUQn1Cgz6
TPU Scarcity and Memory Lag Behind A100s, Prompting Rework
@_The_Prophet__ TPUs had low availability for ages and also low memory relatively on the v6e especially versus the hoppers working pretty much out of the box similar to a100s Grace Blackwell is the next thing that needs reworking so there is...
Grok Imagine Impresses with Quality Hindi Generation
@DevDminGod this is pretty good! i didn't know grok imagine could do such decent hindi
TPUs Offer More Stable Training Than CUDA on Large Batches
@_The_Prophet__ TPUs have been more stable for training than CUDA equivalents for a couple of years now, especially on large batch sizes XLA is pretty good now! For inference it makes even less of a difference (We previously trained sota models on thousands...
Must‑Listen Interview Ignites AI Community, Follow My X Pro Lists
I agree. Incredible interview by @dwarkesh_sp of @ilyasut. I could listen to both for months and not get bored. It's like being at a great university and hearing the best professor. I love X. This just LIT UP the AI community....
