The thesis of Ray Kurzweil's new book is that brain-computer interfaces will expand human intelligence millions-fold, unlocking new modes of thought: this is the Singularity

at Electric Sheep we’re building animal minds that perform real world tasks robustly and at scale.

GenAI vs search. Physics grounding is needed ...

A visualization of sorting algorithms 🔊 [📹 Timo Bingmann]

🚨 Foundational Models for Robotics: Deep Dive 🚨 Foundational models are reshaping AI but can they revolutionize robotics? Let's dive into one of the most hyped topics in robotics - exploring the promise and challenges ahead. Here's what you need to know 👇

Since the amazing Unitree G1 was announced today, I want to reiterate what I think are the two big obstacles to general-purpose robots: - generalizable perception and scene understanding - reliable long-horizon reasoning These are major obstacles for online AI agents as well

If each step of an ai agent is 95% accurate. None of the 30 step work flows will work. Going from 95-> 99.9 is a similar last mile problem as with self driving cars. Easy to hack up a prototype. Hard to make it work reliably at scale.

For anyone curious about nonlinear dynamics and chaos, including students taking a first course in the subject: My lectures are freely available here. m.youtube.com/playlist?list=…

There’s a massive opportunity in automating landscaping, and Electric Sheep is one of the companies pushing the boundaries of embodied AI to elevate landscaping businesses across the US. Michael Laskey, Electric Sheep’s CTO, will share their advancements in End2End learning on

This was a very fun video to make. Shoutout dar for bringing my team and I on to help introduce NEO to the world :) 👋👋

This is such an incredible articulation of the future of humanoids.

Game engine data is way more powerful then you thought. Electric Sheep Our entire perception stack cost only $200 in data synthesis. It enable robots to mow lawns anywhere in the world... with zero maps.

The first Starships to Mars will launch in 2 years when the next Earth-Mars transfer window opens. These will be uncrewed to test the reliability of landing intact on Mars. If those landings go well, then the first crewed flights to Mars will be in 4 years. Flight rate will

this video got me thinking …both the cat and the snake beautiful examples of nature building incredible organisms - hoping someday we can get to robots as intelligent, robust and malleable as this. for organisms - its wild that our compute, sensors and actuators are made from

Apple just announced AirPod features to turn them into hearing aids. Having experienced a harder time talking to my grandparents in their later years due solely to hearing loss, this is such a brilliant idea it’s crazy.

what most people will miss is that o1 is significant precisely because it isn’t an SFT on synthetic data the fact that rl on CoT unconstrained works and doesn’t collapse to gibberish cot steps is really profound more thoughts ⬇️

AGI will be solved in the next 5 years. Not because LLMs or current models are, or will directly lead to, AGI but because today, tens of thousands of the smartest researchers and engineers are working on AI instead of mobile apps and social networks. Hundreds of billions in

The world has a shortage of radiologists, physicians, teachers, software engineers, anime artists, pilots, bus drivers, civil engineers, hotel operators, system administrators, architects, sushi chefs, and the list goes on… The more technology there is, the more people we need.