Robotics engineer @IntrinsicAI + Host of the Audrow Nash Podcast
Lockhart, TX
Joined May 2021
- Tweets4.3K
- Following1.6K
- Followers4K
- Likes10.2K
Pinned Tweet
New episode with Amarjot Singh of @SkylarkLabs!
Why Defense Gets Humanoids Before Your Home Does
We've been hearing how imitation learning won't work in production because it works 70-90% of the time, not the 99.99% needed in most industries. For defense work, like kicking in a door, 70% is fine if it gets a human out of harm's way.
I recommend listening at 1.5x
Timestamps:
3:30 Humanoids that sweep for bombs
17:26 1,000x less data
21:58 Spotting a camouflaged tank
55:01 The humanoid Turing test
1:10:56 Why defense before homes
1:28:21 Advice for founders
Audrow Nash retweeted
I'm all for the Roadster just being awesome, but could the jets also be a safety feature?
This happened to my grandparents recently. Their car couldn't fly out of the way.
(Made with Grok @bot and Blender)
I'm all for the Roadster just being awesome, but could the jets also be a safety feature?
This happened to my grandparents recently. Their car couldn't fly out of the way.
(Made with Grok @bot and Blender)
With AI it's hard not to want to ship every feature I can think of. I'm finding I need a spec to stay disciplined in what I ship.
What do you do when your AI agents working?
I often read the code or dig into things, but then I can only do so much of that and turn to scrolling X or watching YouTube.
I'm not sure I like this. Anyone found something better?
Audrow Nash retweeted
This is why Texas will do better than your state.
There is a center of gravity forming around all the stuff @elonmusk is building between Austin and San Antonio and beyond. It's going to be the closest thing the US has to Shenzhen.
Our hourlong exploration of the tech scene in Texas in full here on X.
AI video editing is actually good now
Everyone’s seen the slopcore, but if you give it your own footage it does a solid job
Jevons paradox: I’m going to start taking a lot more video 😂
New episode with Axel Krieger of Inner Logic!
What stuck with me: surgery is a great place for imitation learning. Thousands of surgeons already train on surgical robots. As Axel put it, what other field has that many expert robot users?
youtu.be/0gRJCzaB7Qo
Hey folks, a quick update: I'm pausing the Audrow Nash Podcast indefinitely—no plans to restart right now.
Life's been tough with our <1-year-old bringing joy (and chaos), my dad's sudden passing, and a messy probate. It's just too much to juggle.
Thanks for all the support and great convos. Hope to reconnect someday! 🙏
What if robots stopped doing just the simple stuff and started tackling the most important jobs on the planet?
I talk with Fred Parietti (@FredParietti), Co-founder and CEO of Multiply Labs (@multiplylabs), about how advanced robotics are reshaping the future of pharmaceutical manufacturing by automating high-value, life-saving tasks.
You'll like this interview if you’re curious about robotics, healthcare automation, or how technology can solve complex real-world problems.
Find it here on X, and on YouTube, Apple Podcasts, and Spotify. Links in comment.
I recommend listening at 1.5x
Timestamps:
0:00 Start
0:53 Multiply Labs Overview
7:17 Pharma Robotics Gap
16:18 Next-Gen Pharma
34:04 Multiply Labs Automation
43:58 Deployment Size
1:00:06 Thoughts on Humanoids
1:10:02 Training Process
1:21:45 High-Value Problems
1:24:58 Key Takeaways
Episode links:
- Fred Parietti's website: federicoparietti.com
- Fred Parietti's X: xitter.cf/FredParietti
- Fred Parietti's LinkedIn: linkedin.com/in/fred-pariett…
- Multiply Labs website: multiplylabs.com
- Multiply Labs X: xitter.cf/multiplylabs
- Multiply Labs LinkedIn: linkedin.com/company/multipl…
Podcast links:
- Podcast Website: audrownashpodcast.com/
- YouTube: youtube.com/@audrow
- Apple Podcasts: podcasts.apple.com/us/podcas…
- Spotify: open.spotify.com/show/74jWpW…
Why do so many robots have legs? Is the humanoid form just hype or does physics demand it?
I talk with Pras Velagapudi, CTO of Agility Robotics (@agilityrobotics), about how real-world engineering and practical constraints led their team to design a two-legged robot that’s built for actual work—not just to look cool.
You'll like this interview if you’re curious about why robot shapes matter, how automation meets the real world, or what it really takes to build a useful robot.
Find it here on X, and on YouTube, Apple Podcasts, and Spotify. Links in comment.
I recommend listening at 1.5x
Timestamps:
0:00 Start
0:42 Agility Overview
5:29 Why Humanoid Design
13:31 Why Legs, Not Wheels
23:19 Digit’s application
38:26 Safety in Humanoids
49:15 Standardizing Safety
1:06:50 Digit in Other Industries
1:15:06 How AI Fits into Robotics
1:29:23 Is Data Still Gold?
1:35:57 The Future of Agility
1:38:16 Key Takeaway
Episode links:
- Pras Velagapudi's website: pras.ai
- Pras Velagapudi on LinkedIn: linkedin.com/in/pvelagapudi
- Agility Robotics website: agilityrobotics.com
- Agility Robotics on X: xitter.cf/agilityrobotics
- Agility Robotics on LinkedIn: linkedin.com/company/agility…
Podcast links:
- Podcast Website: audrownashpodcast.com/
- YouTube: youtube.com/@audrow
- Apple Podcasts: podcasts.apple.com/us/podcas…
- Spotify: open.spotify.com/show/74jWpW…
How does @jonesbrannon from @alley_corp think about humanoids?
🔊 "That's a pretty low bar" 😂
Founder-market fit: Why does no one talk about this?
with @cbames from @BotBuilt_