Shipping robots @BuildMonumental | AI nerd | Ex-Picnic
Amsterdam, The Netherlands
Joined May 2009
- Tweets13.8K
- Following526
- Followers579
- Likes8K
Jelmer Borst retweeted
Sometimes you see a graph that stops you in your tracks
October makes September look like a rounding error 👀
Never seen anything like this in my decade as an investor…
Jelmer Borst retweeted
Human-robot collaboration 🟠👷
Now both the construction site and finished builds look like works of art.
#robotics #physicalAI #bricks #construction
Jelmer Borst retweeted
Here's what this build looked like in Atrium. It's plain magical to write code that becomes a real building.
Being able to vibe code buildings is definitely not far off, if you build the right infrastructure and harness for it.
The model is the easy part
For anyone claiming robotics data is a solved problem or already commoditized:
- Confuses egocentric data with all types of robot-training data
- Has never tested a robotics model zero-shot
- Has never experienced the pain (and joy) of overfitting
- Has never deployed a policy on real hardware, not just in sim
- Has never tested generalization beyond the tasks seen in training or standard benchmarks like robodojo
Robotics data is far from solved.
Huh. So far they were always benchmaxxing. Scores looked great, but using the models were horrible. Maybe they are finally getting it?
This looks like a vibe coded setup. But holy shit.
Last night an AI agent performed brain surgery on a mouse.
We gave it 19 gene therapy vectors in a spreadsheet, a mouse model of epilepsy, access to a motorize surgery rig with cameras and one goal: reduce seizures.
It picked the drug. It picked the brain region. It derived the coordinates from three papers. Then it drove the stereotaxic arms into the brain and injected 250 nanolitres while watching the entry site through the rig camera for bleeding.
A licensed vet gated every irreversible step and held a physical kill switch to stop the AI agent.
Thread on what it decided, what it saw, and the moment it refused to let me override it.
[Writing this on the spot out of excitement, but I've been 18hrs in front of my laptop coding and coordinating this, so keeping it brief and descriptive, more cohesive thoughts on this soon...]
This is when you know they won't ship anything useful any more. PMs and engineers are just bored at this point.
Jelmer Borst retweeted
Our robots built this in the city of Almere: curved walls, different types of bricks and bonds, indented text.
Real world deployment, insane complexity, 100% of every brick done by our robots, with another 15 houses to come in the coming weeks.
People are being silly about this, LOC not being a good measure, stupid PRs etc. But truth is that engs have been shipping shit UIs, and designers are taking the reins - rightfully so.
1. Clearly sees the threat in Instinct etc.
2. Don't see how this will be truly useful without being deeply embedded into Slack etc
We’ve been humbled by the interest in Claude Duet, and are working hard to get people off the waitlist. In the meantime, a few thoughts based on DMs, emails, posts we’ve seen.
→ Who can use it: Pro and Max subscribers in the US, for now.
→ Where: You can use it on iOS, Android, and web once you get off the waitlist. Mac app is in the works.
→ Trust: We designed Claude Duet to be a partner you can trust. More on our safety, security, privacy approach in today’s blog post: claude.com/resources/article…
→ Use cases: We’re starting with work use cases, which is what a lot of people come to Claude for. Professionals, small business owners, founders, anyone juggling a lot. We’re opening the research preview to these segments first so we can co-build together. Over time, we'll support more use cases across work and life.
→ Context: Claude Duet already knows about everything you’ve been working on with Claude. Your chats, memory, projects, and more are available in Claude Duet the moment you open it.
→ Tasks: Tasks are at the center of the experience. Every task gets its own thread you can check in on and steer, so you can multi-task and hand off many things at once. Claude Duet also proactively suggests tasks it can take on, with your approval. We’re glad to see this primitive and dedicated tab resonating. More to come soon.
→ Proactive with judgment: Opus 5.5 is widely loved for complex real-world work and its natural communication style, and we’ve built around it for deep judgment. More work to be done here so it feels like a world-class teammate, coworker, chief of staff.
→ Channels: We went with one agent that works across your main chat and dedicated channels, so each area of work gets its own space. Small business owners and professionals are already using channels for customer support, marketing, bookkeeping, etc.
Starting next week, we’ll share more on how we built Claude Duet and how early testers are using it. We know usage limits and access outside the US are top of mind too, and are working on both. Please let us know what else we can improve - we're reading all your feedback.
AI is continuing to baffle me. It went from autocomplete, to files, to simple PRs, to now just give it super super ambitious stuff.
And it just does it.
The last 2mo have been crazy on how good it is.
This is insane
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
Jelmer Borst retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability.
The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.
But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.
So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.
TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.