Slop is not what we all need.

simonwillison.net/2024/May/8/slo…

Interactive Progress Tracker: Open vs. Proprietary LLMs
huggingface.co/spaces/andrewr…

How to be as 'smart' as Auto-Regressive LLMs:
- memorize lots of problem statements together with recipes on how to solve them.
- to solve a new problem, retrieve the recipe whose problem statement superficially matches the new problem.
- apply the recipe blindly and declare

Meta has more H100 GPUs than everyone else combined??

This is how you win the AI race

Efficient finetuning of Llama 3 with FSDP QDoRA

FSDP QDoRA, a scalable and memory-efficient method to close the gap between parameter efficient finetuning and full finetuning

answer.ai/posts/2024-04-…

DoRA > LoRA

Nice summary from Jonathan Whitaker on HN :D

contact lenses powered by light and tears spectrum.ieee.org/power-smart-co…

Meet Hume’s Empathic Voice Interface (EVI), the first conversational AI with emotional intelligence.

great insights from a16z report on enterprise genAI

- budgets are skyrocketing ($7M '23 -> $18M '24)
- measuring ROI is both art & science
- OpenAI dominates
- open-source is booming -> control

a16z.com/generative-ai-…

#genAI #AI #enterpriseAI

Sanctuary's Phoenix bot, with 'the world's best robot hands,' working totally autonomously at near-human speeds – much faster than Tesla's or Figure's robots

And then there were three...

I got access to the new Anthropic Claude 3 AI a few days ago, so not enough time for a full review, but it was obvious it was GPT-4 class even before they released the testing stats. At the same time, like Gemini Advanced, it doesn't blow GPT-4 away.

AGI today = statistical models that exhibit preferences/biases of a controlling entity

Custom LLM and AI Agents (RAG) On Structured + Unstructured Data - AI Brain For Your Organization

Imagine a ChatGPT-like interface over all your structured (database) and unstructured data. Ideally, you want to ask a question to an AI bot, and it should be able to run multiple

Figure 01 has learned to make coffee ☕️

End-to-end AI system, trained in 10 hours, just by watching humans make coffee

Our neural networks are taking video in, trajectories out

Join us to ship a fleet of AI robots: figure.ai/careers

These are some of the most complex tasks I’ve ever seen a robot do autonomously.

The best part: nothing fancy required.
- imitation learning with ACT or Diffusion Policy
- co-train with static manipulation data

Zipeng's thread has a results summary⬇️
x.com/zipengfu/statu…

So, what new skills does 𝐌𝐨𝐛𝐢𝐥𝐞 𝐀𝐋𝐎𝐇𝐀 unlock when controlled by a neural network?

Check out 𝐌𝐨𝐛𝐢𝐥𝐞 𝐀𝐋𝐎𝐇𝐀🏄 - Learning from Co-lead Zipeng Fu !

x.com/zipengfu/statu…

We open-source all hardware and software of 𝐌𝐨𝐛𝐢𝐥𝐞 𝐀𝐋𝐎𝐇𝐀:

Tutorial: tinyurl.com/mobile-aloha-t…
Github: github.com/MarkFzp/mobile…
Project Website: mobile-aloha.github.io

Introducing 𝐌𝐨𝐛𝐢𝐥𝐞 𝐀𝐋𝐎𝐇𝐀🏄 -- Hardware!
A low-cost, open-source, mobile manipulator.

One of the most high-effort projects in my past 5yrs! Not possible without co-lead Zipeng Fu and Chelsea Finn.

At the end, what's better than cooking yourself a meal with the 🤖🧑‍🍳