Every software engineer hits the same wall: “I don’t know why this broke in prod.” AI coding agents fall apart without the right context. Hud captures how your code behaves in production and surfaces that context in your IDE and to AI coding agents via Hud’s MCP server. MCP… https://t.co/2BN35MPVFA

Valthor is one of Lovable’s earliest team members and a key reason we’ve been able to reliably scale to millions of users. He has a unique ability to take any technical problem and dig into the fundamental components to truly understand the solution. Some people in the team joke… https://t.co/qz91T6Czd3

Our flagship paper on how far careful quantization can really go in practice got accepted as an oral at ACL 2025 (top 8%)! 🥳 Turns out, old-school methods like GPTQ, SmoothQuant, and RTN are quite good when tuned properly. All of the tricks are already in LLM-Compressor! https://t.co/lX1X0XGGeZ

Every billion dollar founder I know dropped off the face of the earth for years. We built the perfect home for these people. It’s called HF0. In our last batch, teams raising after demo day hit an avg valuation of $82M. Applications for F25 open today. https://t.co/LySfXbzBCh

updated open-source object detection model leaderboard 64 checkpoints; key metrics like mAP and F1, including small object detection performance we plan to add latency measurements for T4, A100, and Jetson link: https://t.co/9z9ImuoChZ https://t.co/ff3ipRPSTS

Replace git commands with natural language! Introducing ✨gut✨: A human-in-the-loop agent in the form of a command line tool. 💬 Explain the git actions you want to take in human language 🤖 Wait for the agent to figure out the git command, explain it to you and wait for your… https://t.co/mVkozoQzzR

Introducing the all new supermemory 💥 we killed the AI memory lock-in. A universal memory for all the apps you use: - MCP connect to Claude / Cursor / ChatGPT - Graph view of your life - Projects to organize your life free to use. $9/m for first 100 pro subs. demo below https://t.co/XtYto0rRH2

⭐️ New integration release ⭐️: S3 vector storage! 🤖 Our new S3VectorStore integration combines AWS S3's enterprise-grade scalability with LlamaIndex, giving your agent workflows a robust knowledge foundation that grows with your needs. Build smarter agent workflows that: ➡️… https://t.co/71ADmwp6NF

I found another fun Veo 3 prompt: "Realistic footage of the moon landing, but it took place in [year]" Here is 1883 AD, 1255 AD, 44 AD, 2300 BC, 30,000 BC, and 65 million years ago Yes, there is apparently wind on the moon back then, you are just going to have suspend disbelief https://t.co/7j2NvooWrq

AI doesn't search like humans do. It wants paragraphs, not keywords. It needs 1000 results, not 10. This is why traditional search engines fail AI systems. They were built for lazy humans who can't type full sentences. The future of search isn't about clicks, it's about… https://t.co/tZBep1Bt7b

AI systems don't tell you when they're hallucinating. They just confidently serve garbage to your users. https://t.co/Yz3fMR77rC

Learning without training Google researchers explore the implicit dynamics of in-context learning. "Implicit weight updates from ICL mirror the effect of actual fine-tuning on the same data." This one is more technical but much needed. The findings: https://t.co/BT07FeJaYr

Using expertise from @HamelHusain and @sh_reya 's evals course with good AI coding practices taught in my course means you can use Agentic tools to set up core eval infra fast Automate tests -> view data with custom annotation app -> further analysis in jupyter? Yes plz. https://t.co/bgraEk4OWi

It was missing, so I added @AnthropicAI Opus 4 Thinking and @OpenAI o3 benchmark results to the comparison mix chart 🆚🔎 Vibe check pending, but on benchmarks it seems that we got an open model competitive with Opus 4 / o3 / Gemini 2.5 🤯 https://t.co/76XnKBB5sK

Google just discovered a powerful emergent capability in Veo 3 - visually annotate your instructions on the start frame, and Veo just does it for you! Instead of iterating endlessly on the perfect prompt, defining complex spatial relationships in words, you can just draw it out… https://t.co/DWsxiVGBuq

Tired: Prompting Veo 3 videos with JSON. Wired: Prompting Veo 3 videos with PowerPoint. https://t.co/9HUQLXlPzz

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://t.co/PGAiVGnlB9

Build an AI agent that can generate visual reports 🧑‍🔬🤖 It’s fine to have AI agents generate text, but it’s way more fun to have them generate a full multimodal report with interleaving text and images. To do this you need a combination of fine-grained data extraction combined… https://t.co/cQ2pPZD4Ba

Pull Request is up for testing and review. There's still a lot to be done, but it's in a functioning state now. https://t.co/34yoIVsGff

Wonder if the course is right for you? 1. Have you trained a model? (If yes, you fulfill the prereqs) 2. Do you know nothing about distributed training? (If yes, we’ll help you get there) 3. Do you train models daily on clusters, and want to know the latest techniques? (If yes,… https://t.co/2VA6WkNHvF

Very sad to learn of the death on 18th July of Margaret (Maggie) Boden, a titan of cognitive science and AI. I met her many times, and respected her greatly. https://t.co/LA45gP7ncx https://t.co/pVdlEckemS

🖊️New blog post: We're breaking down DSPy's few-shot optimizers (LabeledFewShot, BootstrapFewShot, BootstrapFewShotWithRandomSearch). While we're at it, we'll also explore "traces", "bootstrapping" and when to use each optimizer without the jargon. (link in comment cuz algo) https://t.co/hOceFoXXDJ

Claude Code sub-agents is probably the easiest way to build complex, custom agentic systems. Not just for code, but for anything you can imagine. Watch me build a multi-agent deep research system with Claude Code subagents. Chain subagents with /commands for reliability. https://t.co/8zGNPMSSfV

Anthropic's recent blog post provides a good summary of some of the recent work applying mechanistic interpretability to biology. I am very bullish on this research direction! https://t.co/g2UdvcDbsI

If you want ONE place to keep up with AI coding agents, you should pay attention to what @isaac_flath and @intellectronica are putting together: https://t.co/meoZvnNzUQ. I've worked with both, and they have phenomenal taste. Isaac: 7+ years working on dev tools in both open… https://t.co/WQISDjldzh

Using novel games to test AI makes a lot of sense and the ideas remind me of @pippinbarr’s chess variants. Playable here: https://t.co/dxoW8vY5dq https://t.co/rBJtNauaMW

+1 more

To what extent can you automate or delegate evals? Is there a way to make it "not your problem?" 😅 Part 1 of 1: You should absolutely automate parts of it as long as a human is in the loop. Many people are a bit too aggressive here, so you have to be careful, some guidelines… https://t.co/K5fcFKS6zl

did anthropic vibe-code their infra? https://t.co/4w9N0cXSj1

"Vibe checks" are great—until you need to scale. In this clip, @HamelHusain and @sh_reya break down why relying on human intuition isn’t enough when it comes to evaluating product or model quality at scale. Instead, they explain how to codify those gut checks into scalable,… https://t.co/d70rKzOWdC

✨EVERY SHOT WAN TAKE ✨ WAN 2.2 cinematic creativity open source release Live Broadcast Time: Jul 28th, 2025, 20:00 (UTC +8) https://t.co/9Dc28AYKb8

GLM-4.5 looks like a big deal! > MoE Architecture > Hybrid reasoning models > 355B total (32B active) > GQA + partial RoPE > Multi-Token Prediction > Muon Optimizer + QK-Norm > 22T-token training corpus > Slime RL Infrastructure > Native tool use Here's all you need to know: https://t.co/QXFr4QBxkk

Just upgraded the FAQ even more, now has 1. 📺 Embedded videos and diagrams 2. 🔍 "Focus view" which places each FAQ on its own page for easier sharing / less distraction 3.🔉 Listen to audio version (customized podcast narrative) 4. 📑 Download PDF version directly from the… https://t.co/yBNshE88sC