Arash Tajik (@atajik) header
Arash Tajik (@atajik) profile picture

Design In, Product Out | Making every factory an autonomous factory | Founder, Axomind, ex-Apple

San Francisco, CA
Joined December 2010
Photo by Arash Tajik (@atajik)Photo by Arash Tajik (@atajik)Photo by Arash Tajik (@atajik)Photo by Arash Tajik (@atajik)Photo by Arash Tajik (@atajik)Photo by Arash Tajik (@atajik)
I am gonna quote @friedberg verbatim, because that's the most important perspective in AI: "I think the biggest opportunity for AI is on the frontier. That is where basically businesses and people can do things that humans and businesses have never been able to do before. And that means crafting solutions to really difficult engineering problems, to really difficult scientific problems, to previously untractable problems. And that's where I think that the frontier gets expanded by 100x or a thousandx versus AI being some like iteration on existing business processes to save money and make things more efficient." youtu.be/G5w7EyF-aFQ?t=1889
68
“we can learn a lot about where the world is going by tracking where the smartest young people are going”
riemann hypothesis solved by levent alpoge at anthropic it's the single most important math problem, at least 1000+ theorems build on it fun fact about levent alpoge from his friend in 2010, as a freshman at harvard, he was already considered one of the two most talented mathematicians in his class he once told his friend that riemann hypothesis is 5x harder to solve than poincare hypothesis, the biggest problem solved pre-ai by 2024, ai was getting imo level good (high school math olympiad) alpoge was disillusioned, and said working on math (the traditional way) is pointless, and left harvard to go to anthropic that year we can learn a lot about where the world is going by tracking where the smartest young people are going
1
120
Humanoids will enable the recursive self-improvement (RSI) in the physical world. Generalized brain on a generalized form-factor is the requirement to build systems that build systems.
107
Robotics has the most to lose if Jensen leaves NVIDIA. Few leaders have given this community as much as Jensen and NVIDIA's talented team have.
2
230
The world of atoms is only an actuator away from the world of bits.
We're at the threshold of acceleration in the world of bits being transduced to the world of atoms
1
1
307
It is quite unsettling to see American's view of AI is so low and is sitting along with countries like Afghanistan and Malawi. We need to be better at communicating the benefits of AI.
2
255
Ken Griffin Nailed it: If you want to reindustrialize America you need robotics. People don’t want to live that dream of working at a factory anymore! We need robotics to run our factories.
Citadel founder Ken Griffin warns President Trump faces serious challenges ahead of the midterms, arguing that his unpopular Iran war, “disgraceful” treatment of illegal immigrants, and inflationary tariffs are alienating Americans and making life more expensive. 🇺🇸 “The president’s going to have some challenges here. There are parts of his policy that have been very unattractive to the American people. For example, the war in Iran is very unpopular. I think he’s done a very poor job explaining why. I think it’s important that we wake up in a world where Iran does not have a nuclear weapon. That’s existentially important to our country. The president has not made that case in a powerful, meaningful way to the American people. I think the way that we deal with illegal immigrants is disgraceful. If you came to our country when our borders were open, you’re gainfully employed, and you’re working to find a better life for your family, there needs to be a way to stay in our country. It may not be a path to citizenship, but it’s a path to being a contributing member of our society. We need to grow our economy. If we’re going to make good on Social Security, Medicare, Medicaid, and other entitlements, hardworking people are a big part of how we’re going to make that happen. The tariff battles with Canada, with Europe, with our allies, have just been a disaster. I talked about this when he first started this in his presidency. To be blunt, I hate to say it, but I’ve been right. It’s been pro-inflation. It has taken supply chains around the world and severed them. If we want to reindustrialize America, we should really be embracing robotics. One of the reasons I gave a recent gift of $3 billion to Carnegie Mellon University is their leadership in robotics. We want a reindustrialized America with robotics running our factories. That’s forward-thinking. That’s how our country secures greater independence from the rest of the world on items important to our national interest.”
1
162
What happened to SW engineering last year will come to HW engineering next year. Exciting times.
ChatGPT and Codex are helping engineers build for the physical world. We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
1
1
244
This important
$20B valuation at a typical 20x revenue multiple means about $1B a year in revenue (hypothetically). Robot brain APIs are launching at about $3 per robot hour, that's roughly 167k robots working full time shifts on @physical_int's model. For context, about 600k industrial robots were installed worldwide in all of 2025
168
Very much agree with this analysis. #10 (deployment) should be top of the list. That's where economically useful robotics happen.
10 Spicy Takes After Talking to 50 Friends at the Frontier of Robotics 🤖🔥 1/ Astra IS robotics' GPT-3 moment. Ironically, now that it's finally here, some in robotics seem reluctant to acknowledge it. Perhaps because it (sadly) didn't come from a robotics company? 2/ Robotics is becoming just another vertical for frontier labs. Not so different from coding in some respects. Why reinvent general intelligence for every industry? 3/ Robotics was too hard to underwrite until now. No convincing pretraining scaling law, no convergence on models, data, or hardware. No wonder some critics dismissed it as a fragmented industry playing an expensive "demo game". Astra may finally offer a credible path to scale. 4/ Robotics foundation-model startups have a case. But how strong is it? The argument: Astra handles System 2 (reasoning), while specialized models retain System 1 (skills) and System 0 (motor control). Fair. But isn't that already a narrower value proposition? And who says the boundary stops there? 5/ Follow the talent. Word is OpenAI reorganized its 30–40-person robotics team around the core Astra model. Hearing similar moves toward general models at @Tesla_Optimus. Org charts may tell us more than any demo. 6/ The robotics data gold rush is real. Million-hour procurement programs, some potentially reaching tens of millions of hours. Talk of budgets approaching $1B per frontier lab. 7/ Bullish on robotics hardware volumes. As intelligence becomes more accessible, the barriers to building useful robots could fall dramatically. The moat shifts toward reliability, dexterity, integration, and deployment. 8/ Robotics could fuel the next semiconductor and memory supercycle. A 100B-parameter model running on the edge requires ~50GB just for 4-bit weights. Now multiply that by millions of robots. Why do you think @JensenHuang is so obsessed with robotics? @elonmusk saw it coming: xitter.cf/elonmusk/status/211477… 9/ Is self-driving immune to the general-model threat? After coding and robotics, why assume self-driving remains the exclusive domain of proprietary, vertical models? Robotaxi increasingly feels like a solved problem to me. But if and when frontier labs move into self-driving, we could see an Apple vs. Android dynamic. Think about the implications for $TSLA, Waymo, and other OEMs. 10/ The next bottleneck may not be intelligence. It's deployment. Models are improving faster than our ability to deploy robots in the messy real world. Integration, hardware variance, failure recovery, and economics remain brutally hard. The real moat may be turning intelligence into reliable, profitable physical work. -- I realize some of these takes may rub people the wrong way, especially those who have spent years doing the hard work of advancing robotics. That's not my intention. I'm sharing them to spark debate, not to declare winners and losers. Astra has only been out for a month or so, and it's far too early to draw definitive conclusions. More importantly, when the TAM is this enormous - potentially the largest in tech - there should be plenty of room for different approaches and many winners. ++ Full diligence report: robonomics.substack.com/p/ro…
1
293
ahem...The deepest change in robotics takes the from of code!
Talking to Amjad yesterday, I realized Replit is more important than I'd previously thought. It's the way AI pervades organizations. Organizations can become somewhat aligned with AI by using AI-powered apps, but the deepest change takes the form of code.
210
Robots don't get a minimum wage. If your robot is just an easy drop-in, on wheels or legs, you're not building a moat; you're entering a race to the bottom. Easy to drop in = easy to swap out!
1
148
I'll be at FABTECH in Las Vegas this week, Oct 21 to 23. If you run a fab or build equipment for one, I'd love to meet at the show. At Axomind, we turn CAD into work instructions for robots, so the factory keeps up every time the design changes. We're building this for metal fabs: sheet metals processing, welding, and assembly.
1
1
134
Congrats to the Arda team. This is what we've been building at Axomind since day one: turning CAD into work instructions that program robots, so the factory adapts when the design changes. That's the foundation for autonomous factories. Design in, Product out.
We’ve been talking about how the industrial wave is coming. With @Arda_Industries, it’s officially here. Today, Arda comes out of stealth with an AI agent that turns CAD files into assembly instructions. It’s already been deployed across every division at @anduriltech, reducing engineering workflows from 30 hours to 2.5. That's step one of a three-act plan. Next: use those instructions to train robotic production lines that adapt quickly to new designs. Then: make it possible to build anything domestically. Thrilled that we at @Felicis have joined this round. More at @felicis: felicis.link/wJRN7I3
1
111
Likewise at Axomind. And I believe it’s so amazing that a 3-person team is now pushing more code than what it used to take a 20-person team!
Claude writes about 90% of Claude Code's code. From June to September, merged PRs more than doubled while the share reverted fell by more than 40%. We wrote about how we changed code review to get there, and how to adopt it: claude.dev/blog/solving-code…
63
Well done! Touch is key in dextrous manipulation!
What can a robot do without vision? We gave our robot a sense of touch and showed it can recognize and retrieve objects with its cameras blindfolded. Introducing Rhoda’s new tactile sensor—it feels pressure, texture, vibration, slip, and temperature, just like your fingertips.
1
132
Gemini 4 is quite good at robotics. As models like Gemini and GPT learn the physical world, it raises the question: where is the line between digital and physical AI? imo any task a human expert (in this case a robotics or automation engineer) can do on a computer, AI models will do. Whether the manifestation is on a screen or in the physical world won’t matter.
Gemini 4 Argon taught a robot to play cornhole just by watching a video. It mapped the physics, coded its own simulation, and taught @Apptronik's Apollo humanoid to sink 7 throws in a row.
111
on moats...
Moats! Does anyone have one anymore? Yes. Yes, they do. But only a small subset of what people call moats create real long-term value and defensibility. Not actually a moat: 1. Distribution: This does matter more than ever. Building products has gotten easier, but getting them to customers has not. Distribution, however, does not create durable defensibility. It helps create the raw ingredients for one, but it is not itself a moat. 2. Execution: Yes, speed is critical for startups. But a company that has to sprint forever is not a good business. Ultimately, it will misstep, slow down, and competition will catch up. Speed gives a company more shots on goal as it tries to build a great product and a great business, but it must eventually find something more durable. 3. Brand: The idea is that when everyone is building the same thing, the most resonant brand will win. But for startups, brand works more like customer acquisition spend: Put money in, get growth out. It’s not something that will effectively keep others from competing long-term. Brand can become a true moat over time (as with Coke or Nike or Apple), but it just takes too long to really matter for most startups. What you're looking for is compounding advantages: a self-reinforcing flywheel that gets stronger and stronger over time, ultimately making a company very hard to compete with long-term. There are only four compounding advantages that really matter. Every major tech company has at least one: 1. Switching costs: The more deeply embedded a product is in the way a customer works, the harder it is for that customer to switch. This increases retention and revenue per customer, which can be reinvested into the product. 2. Network effects: The more users a product has, the better it gets for everyone. This increases engagement and attracts more users. 3. Scale economies: The more volume a company has, the larger and more efficient its operations can be. This lowers unit costs, funding lower prices or a better product at the same price, further driving up volume. 4. Data effects: The more usage a product has, the more data it generates. This data can be used to improve the product and increase usage. If you can’t see the potential for any in the startup you’re considering joining, that’s a red flag. In today’s incredible post, @danhockenmaier shares a simple framework you can run in about an hour after getting an offer: 1. Identify a startup’s compounding advantage 2. Gauge how strong it could get 3. Measure whether it’s working Inside: dozens of real company examples, the specific metrics to track, and benchmarks for what good and great look like. Read it here: lennysnewsletter.com/p/how-t…
3
129