We Are Volume’s Post

Four ways generative AI is boosting robotics research: - Bridging the sim-to-real gap - Bridging the human-to-robot gap - Bridging the robot-to-robot gap - Coding better reward algorithms for robot control models https://lnkd.in/enDNj4fJ

🔙 Throwback to DevTalks Romania when Professor Sethu Vijayakumar presented on 𝐌𝐚𝐢𝐧 𝐒𝐭𝐚𝐠𝐞 an impressive session about AI, Robotics and how the Future will be in this world. 👀 Take a look on this article and watch his full session here:

Robotics has been in a bit of a winter for several years now because the pace of innovation has been (relatively) slower(er) as compared to software. That's about to change as the advances in generative AI and it's code generation capabilities come to robotics. Companies and universities alike are seizing the moment . Below is just a small example of what is to come. Very excited to see what comes next for Pittsburgh as well as the robotics capital of the world https://robopgh.org/ https://lnkd.in/gN7QZKDe

A must-read for ML researchers and engineers. Robotics is a domain that keeps AI/ML research honest. If an AI/ML approach from your paper is applicable to robotics in principle but can't be deployed on a robot w/o "future work" 😉, keep on trying! This is also a reminder to self.

An excellent read. Billlion$ spent on the easier parts of ML/AI and robotics is like spending all your $ to get a rocket to kiss the Kármán line — *reliably and safely* delivering things (including people!) to orbit and beyond is exponentially harder and doesn’t translate linearly from the easier stuff. The eternal adage that the last 10% of a problem takes 10x the time and money of the 1st 90% applies. VC - and capital formation in general - have a serious issue in fetishizing the funding of 100 competitors doing “first 90%” stuff, with increasing amounts poured into the most self-congratulatory winner of the “race to easy”. Companies that legitimately deserve credit for addressing that last 10%, like Waymo making non-murdery autonomous vehicles, or Agility Robotics delivering humanoid robots that actually work safely beside people (instead of delivering hacked up demos), or folks that are making AI models that don’t deliver stochastic and harmful hallucinations, like SpaceX and Rocket Lab consistently delivering things to orbit besides press releases, are the bellwethers for “getting beyond easy”. The same small roster of companies going the whole distance for societally-relevant outcomes (with AI/ML & real-world proven physical automation systems) can also be found in climate tech delivering actual solutions versus just winning grants for wishful thinking, and in biotech delivering life saving drugs to actual patients (eg, Creyon Bio in 7 months and under $2m) versus demoware. The capital should be doing there, not to the boys high-fiving themselves 20 yards from the goal line.

This foundation model research could point to a leap in adaptability and the ability to tackle problems not explicitly covered in training data, almost like a robot with… human-like intelligence! - - #MachineLearning #Robotics #AIoT #Careers #DataScience #ArtificialIntelligence #AI

If you want a robot to make you a cup of coffee, Figure AI has you covered. In a nifty demo reel, the newly multi-million-dollar-funded robotics darling showed how its Figure 01 robot could successfully brew a Keurig cup of joe. It's not impressive, you could do it faster and at least hand me the cup but Figure 01's ambition, to bring a general-purpose humanoid to life, well, that's worth some consideration. https://lnkd.in/e4MWWVen

Predictions for robotics in 2024. Looks like we've read our crystal ball correctly, we have been diving head-first into RUST for development which has been especially good for our aerospace and robotics work, we are seeing more interest in AMRs outside of warehousing, and we've invested a lot of time in generative AI and emerging AI technologies like VQAs and LLMs. https://lnkd.in/ghxVfydF

New robotics research from Meta AI and Carnegie Mellon University Robotics — RoboAgent can acquire a wide diversity of non-trivial skills and generalize them to hundreds of diverse unseen scenarios — all with an order of magnitude less data than prior works in this space. Additionally we're open-sourcing RoboSet, a large, high-quality robotics dataset collected with commodity hardware, to support and accelerate open source research in robot learning. More details ➡️ https://bit.ly/3qtP04r Get the dataset ➡️ https://bit.ly/3QFKDh5 RoboAgent can exhibit a diverse set of 12 non-trivial manipulation skills across 38 tasks and can generalize them to hundreds of diverse unseen scenarios — beyond that, it can also evolve its capabilities with new experiences.

#ICYMI Gill Pratt, CEO of Toyota Research Institute, is joining Marc Raibert on the TechCrunch Disrupt stage for a fireside panel hosted by Brian Heater on September 21. Come get the latest insights on developments in #robotics and how #AI could drive future advancements in the field from two industry leaders! Learn more and get tickets at: https://lnkd.in/ePiZxBq4   #ArtificialIntelligence #Future #TCDisrupt2023