robotics
Machines that act in the world: humanoids, training data for robots, and how far behind language models they run.
FilterEveryone, all time115 from 40 people
- FC
François Chollet quoted
Creator of the Keras deep-learning library and the ARC-AGI benchmark
Our readingActual autonomous robotics requires engineering causality and discrete reasoning back into the loop.
Their wordsIf we want actual autonomous robotics we have to engineer causality and discrete reasoning back into the loop.
- 2 weeks earlier
- MG
Molly Graham quoted
Co-founder of Glue Club and The Guild
Their wordsone of my biggest realizations was that delegating something to a robot is not the same as giving it to a human because you cannot get rid of the oversight, right?
↗Molly Graham: The grief, burnout, and opportunity hiding inside the AI transitionyoutube.com 12th of 20 in this recording
- MG
Molly Graham quoted
Their wordsLike I think there are some Legos that shouldn't be given away. um that the robots like both like don't deserve and like that we should hold on to.
↗Molly Graham: The grief, burnout, and opportunity hiding inside the AI transitionyoutube.com 16th of 20 in this recording
- MG
Molly Graham quoted
Their wordsAnd I think things that require trust like that are you know fundamental to relationships like we should not be outsourcing these things to robots
↗Molly Graham: The grief, burnout, and opportunity hiding inside the AI transitionyoutube.com 17th of 20 in this recording
- MG
Molly Graham quoted
Their wordsYou individually are phenomenal at a set of things. Like do not outsource that. Like you can enable it, you can amplify it to use your data, but don't outsource it to these weird robots that you know are effectively summer interns.
↗Molly Graham: The grief, burnout, and opportunity hiding inside the AI transitionyoutube.com 18th of 20 in this recording
- 1 day earlier
- NS
Noah Smith quoted
Economics writer; author of the Noahpinion newsletter
Their words"robots escape their prisons and colonize the internet" is now going to be the story of the internet for the foreseeable future.
- 3 days earlier
- MS
M.G. Siegler quoted
Technology writer of Spyglass
Their wordsThe AI is better now, but there are still too many things in the real world that sort of require humans in some capacity – at least until the actual robots come.
- 3 days earlier
- TE
Terence Eden quoted
British open-standards and open-source technologist
Their wordsThe fact that we don't have swarms of drones prefabbing houses in the relatively benign atmosphere of our planet should be evidence that maybe these claims aren't quite matched with reality.
↗Book Review: How to Build a Space Station by Jonathan Morrison ★★★⯪☆shkspr.mobi 2nd of 3 in this piece
- 5 days earlier
- ML
Michael Lopp quoted
Engineering manager who has worked at Apple, Netscape
Their wordsI do know that the robot's primary job in responding is to make sure you get a satisfying answer. I didn't say correct; I said satisfying.
↗Robots Have No Voicerandsinrepose.com 1st of 2 in this piece
- 1 day earlier
- KK
Kevin Kelly quoted
Founding executive editor of Wired and author of What…
Our readingRobots will not be able to beat the best human basketball players in a full game by 2040.
Their wordsI bet against this timeline and say they will not. The winner has to buy a round of gelato.
- KK
Kevin Kelly quoted
Their wordsThe greater lesson I think we will learn is that a general humanoid robot is not going to be that useful.
- KK
Kevin Kelly quoted
Their wordsBasketball robots are likely to invent new basketball moves that humans didn't think of doing, or couldn't do easily.
- 5 days earlier
- EM
Elon Musk quoted
Founded SpaceX in 2002
Their wordsAI+robots will more than double the global economy in less than 10 years
- 5 days earlier
- FL
Fei-Fei Li quoted
Sequoia Professor of Computer Science at Stanford
Our readingThe biggest problem in robotics today is data, not chips; chips become the bottleneck later.
Their wordsWe should zoom out and recognize the biggest problem right now in robotics is actually data. One day it'll be chips, but for now it's data.
↗Why Fei-Fei Li Is Betting on Spatial Intelligenceyoutube.com 8th of 13 in this recording
- FL
Fei-Fei Li quoted
Their wordsThere is also a very important step called randomization. Is that you have to take the same environment and then randomize the conditions. So, the cable doesn't literally only, you know, uh bend this way. It can bend a different way or the box can have different sizes, colors, different lids, and all that.
↗Why Fei-Fei Li Is Betting on Spatial Intelligenceyoutube.com 9th of 13 in this recording
- FL
Fei-Fei Li quoted
Their wordscuz we don't yet have a a frontier foundation model that's robust enough for for uh robotics.
↗Why Fei-Fei Li Is Betting on Spatial Intelligenceyoutube.com 10th of 13 in this recording
- 2 days earlier
- ER
Evil Rabbit quoted
Founding designer at Vercel
Their wordsthe agent by itself you know if it doesn't have the human factor involved it will not do anything that at least not at this point maybe in in the future maybe when AI and robots get more sophisticated but at this point we are directors we are creative directors
↗Evil Rabbit - The story of Evil Rabbit and designing Vercelyoutube.com 8th of 15 in this recording
- 13 days earlier
- EM
Elon Musk quoted
Our readingHumanoid robots paired with AI will eventually deliver medical care to everyone on Earth.
Their wordsOne day, Optimus + Grok will provide incredible medical care to all the people of Earth
- 1 day earlier
- ML
Michael Lopp quoted
Our readingVery soon, an AI assistant will be a constant presence, virtual or physical, with nearly everyone.
Their wordsI believe I see where this is all going, and very, very soon, there will be a robot (which happens to have artificial intelligence) sitting virtually or physically near me.
- ML
Michael Lopp quoted
Their wordsThe same skills I use to lead the humans, I use to lead the robots.
- 4 days earlier
- IT
Isaiah Taylor quoted
Founder and CEO of Valar Atomics
Their wordsIf you keep asking this question, what you realize is energy is the fundamental input. When we figure out AI and robotics that allows us to do semi-aututonomous manufacturing, energy will become the cost of all things, right? The cost of buying a thing will become the cost of energy used to make it.
↗How Nuclear Will Unlock Energy Abundance with Valar Atomics Founder Isaiah Tayloryoutube.com 18th of 18 in this recording
- 1 day earlier
- TK
Travis Kalanick quoted
Co-founder and former chief executive of Uber
Their wordsIf you're doing highcale tasks, industrial scale tasks, you are going to make specialized machines that move and act in the physical world. The key part is move. If they're moving, they got to have wheels.
↗Travis Kalanick: How AI Will Transform the Physical Worldyoutube.com 4th of 14 in this recording
- 2 days earlier
- CF
Chelsea Finn quoted
Assistant professor of computer science at Stanford
Their wordsUh and this means that they're going to be far more useful when they're operating fully autonomously. And as a result, this requires us to develop physical AI systems that make far fewer mistakes than the machine learning systems that have been deployed thus far.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 2nd of 22 in this recording
- CF
Chelsea Finn quoted
Their wordsUm, and while this generally improves the reliability of the model, uh, people eventually get tired and it's hard to get really, really high reliability with a person that's manually tuning this. And so what would be even better is if the AI system itself can iterate on the scenario in which you want it to have higher reliability where it on its own automatically seeks out places where it needs more data, where it needs more supervision.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 3rd of 22 in this recording
- CF
Chelsea Finn quoted
Their wordsNow maybe this isn't completely out of the question but this would be quite challenging uh to do and that's because the calculus is a little bit different. We're not just running compute to optimize for a use case. We're actually running the robot in the real world and using the hardware and attempting the task in the real world.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 4th of 22 in this recording
- CF
Chelsea Finn quoted
Their wordsand trying to fold two boxes together isn't useful data that will teach the model how to get better at the task. And so that would be kind of wasting a lot of time on the robot attempting to go down the wrong path for solving the problem. And so instead of spending a lot of time trying to do that task, what we'll do is we'll actually have a human intervene and show the robot what to do and how to recover from that situation.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 5th of 22 in this recording
- CF
Chelsea Finn quoted
Their wordskind of going back to this reliability question, we took this policy and we ran it not just once, but we ran it for 13 hours straight. Uh and we basically wanted to evaluate is this policy not only good at making a latte once, but can it do so reliably to the extent that it would be needed to be useful in the real world?
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 6th of 22 in this recording
- CF
Chelsea Finn quoted
Their wordsSo, you might be surprised to hear that most state-of-the-art foundation models for robotics have no memory or no context. They're just operating on the current sensor observations, the current camera readings, uh, and predicting actions based off of that.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 7th of 22 in this recording
- CF
Chelsea Finn quoted
Their wordsAnd then for longer memory, for memory that spans multiple minutes or multiple hours, we don't necessarily need video of exactly what happened in that past uh in that past history.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 8th of 22 in this recording
- CF
Chelsea Finn quoted
scaling lawsreinforcement learning
Their wordswe see that the across the board the single PIO like pre-trained PIO7 model matches or outperforms the fine-tuned specialists that were developed with reinforcement learning post-training for those downstream tasks.
↗Chelsea Finn: This is the State of the Art in Roboticsyoutube.com 11th of 22 in this recording