Smashing idols and pretty lies. Hammer. Gadfly. Parrhesiast.
Milford, CT
Joined June 2026
- Tweets1.2K
- Following1.4K
- Followers105
- Likes2.2K
Robert Zapp retweeted
Sam Altman often says that AI will lead to a “new Renaissance of creativity and discovery”. Patrick Collison and Tyler Cowen ask why the 21st century has no aesthetic of its own, citing the Bauhaus as the 20th century exemplar.
Know how the Bauhaus started? In 1907, electrification was beginning to transform life in Europe. AEG, the largest electrical company in Germany at the time, hired the architect Peter Behrens as its artistic consultant. He overhauled their brand identity, revolutionised their product design and designed the AEG Turbinenfabrik in Berlin, a vast factory for steam turbines, essentially the data center of its day. Today it’s considered one of the foundational monuments of what we think of as “modern” design. Behrens ran a large studio full of talented young assistants during this period. Among them were Le Corbusier, Mies van der Rohe and Walter Gropius, the architect who founded the Bauhaus ten years later. Meanwhile the creative and cultural divisions of large tech companies today are full to the brim with corporate middle-managers.
Behrens and the Bauhaus believed that beauty vs utility was a false dichotomy, that technology and art formed part of a greater project of progress. The tech industry today has an opportunity to take on this mantle and reimagine how the world looks and feels, not just how it works. But for all the professed love of risk in Silicon Valley, they seem too conservative to try.
Robert Zapp retweeted
A University of Chicago psychologist published a short paper in 1984 proving that the average tutored student outperformed 98% of students in a normal class, and almost nobody outside education research has read it.
His name was Benjamin Bloom.
He called it the 2 Sigma Problem.
Two of his PhD students, Joanne Anania and Joseph Arthur Burke, compared three ways of teaching the same material:
> A normal class with one teacher and a room full of students.
> Mastery learning, where students take regular quick tests and get corrections before moving on.
> One-to-one tutoring, combined with mastery learning.
The tutored students beat the normal class by two standard deviations.
In Bloom's words, "the average tutored student was above 98% of the students in the control class."
And about 90% of tutored students reached a level that only the top 20% of the normal class reached.
Here's the part that matters.
These were ordinary students in ordinary schools. Same kind of kids, different conditions.
Bloom's conclusion was blunt. One-to-one tutoring works, but it is too costly for most societies to provide at scale. So he set a challenge for researchers: find methods of group teaching that work as well as a personal tutor.
He also ranked the cheaper levers. Feedback and correction before moving on was worth about one standard deviation by itself. Graded homework, about 0.8. Cooperative learning, about 0.8.
For 40 years, that challenge mostly sat there.
Now AI tutors are being pitched as the answer.
But the early evidence says the design decides everything. A tutor that hands you answers is a shortcut. A tutor that makes you try, checks you, and corrects you is closer to what Bloom measured.
The 2 sigma never came from the tutor knowing more. It came from someone making sure you could not skip the hard part.
Who was the one person who taught you one-to-one and changed how you learn?
Robert Zapp retweeted
Outside the elite pipeline of $200k+ jobs in NYC straight out of undergrad (sub 1% acceptance typically),
the people who I know doing best financially and finding the most fulfillment have largely taken an entrepreneurial path.
They’ve built something or figured out how to make themselves useful without waiting for someone to hire them.
You can feel the ladder people were taught to climb being pulled out from underneath them.
My suspicion is that this goes much deeper than a bad job market, as a reorganization of the economy around the nature of work and the firm itself happens arounds us.
The top jobs demanding exceptional judgment, intelligence, and initiative probably have more runway.
But if transformative AI delivers on its promise, I don’t think any industry stays recognizable indefinitely.
If you’re young, the response should be with courage, initiative, and creativity. Learn the tools, experiment, build things.
A degree is no indicator for the ability to create value under changing conditions.
In some sense it is unsettling to inherit a world where the familiar paths feel less dependable.
But we’re also getting tools that make previously impossible ambitions accessible to a single person.
This is such a bleak story about what it's like to try to land a job right now. nytimes.com/interactive/2026…
Robert Zapp retweeted
the old education system is like dead dead
there's literally never been a better time in history to homeschool your kids with AI
you can now just prompt it to create an entire curriculum for your kid, give lessons, generate interactive quizzes to see what they're actually learning, explain things in a hundred different ways until they understand, adapt the difficulty and pace to their progress, generate videos, etc etc
you could probably give your kids a much better education in a fraction of the time they'd otherwise spend sitting in a classroom learning the same things at the same pace as 30 other kids
and then spend all that extra time playing with them, exploring, doing sports, reading, traveling, whatever
it's actually insane when you think about it
Physics is hard to visualize through static images alone, so I turned my 174 pages of high school physics notes into an interactive website.
It covers mechanics, electricity, and magnetism, breaking down complex concepts into intuitive visualizations, real-world scenarios, and interactive problems.
Check it out here: atlas-of-physics.vercel.app/
Robert Zapp retweeted
We have discovered a substantial number of vulnerabilities in the German legal system using LLMs.
These may enable access to confidential documents, travel without an otherwise required visa, enforceable leave for soldiers at NATO’s JSEC, and more.
Similar to computer systems, the written law can contain vulnerabilities that invite exploitation. In our blog post, we share a selection of vulnerabilities, and present potential exploits for each. To our knowledge, none of these have been previously recorded. We share only those vulnerabilities whose exploitation we believe would not cause serious harm.
blog post: felixringe.com/legal-exploit…
Robert Zapp retweeted
Palantir CEO Alex Karp says the AI labs want you to buy into a future where you:
"Have no job, adversaries win, and everybody who does win is a small, tiny group of people living in a tiny place that believe because they eat vegetables and don't support war fighters, they deserve to have the total means of production of this country—and the rest of us should just sit back and absorb the cost of that revolution, which we're paying for."
Anthropic CEO Dario Amoedi and his sister Daniella Amoedi are both vegetarian:
“It seemed convincing to me.”
Via @WalterIsaacson @kscope_nyc
Robert Zapp retweeted
Introducing *Switch*: a *harness framework* for *decision models* like Jev
I looked for this the other day and was surprised to not find anything like it
You obviously cannot use a decision model like an LLM in a coding agent harness like @pidotdev, because Jev cannot generate tool calls
But what if there was a framework that "generated the tool calls" for it programmatically, and Jev only had to pick out of that finite set of choices
That is what switch does. You pick a "domain" to implement the decision model in, and it provides a core API to define your state, and the functions that will derive decision model questions out of that state. Chess is a domain, as well as being a rubric judge for evals
It basically standardizes what I imagine a lot of people are already vibe-coding, to an abstract interface: calling a decision-model API through multiple steps, i.e. an agent loop over finite choices
Because of the finiteness of the output of a decision model, you cannot have a general pure decision model harness. Instead, you need to implement specific ones that generate the choices programmatically (or have an LLM generate the choices, but I find the former more interesting)
A primary use case I have for this is use as a judge in an eval. For example, At the end of a task run that involves a bunch of documents, most people currently follow this approach: extract text out of PDF/docx/md files, and dump them right into the context
But what if the visual formatting of the output is a judging criterion? Then you have to dump an image of every page in the input state. But what if you have 200 pages, and your question does not need looking at all of them? What if Jev could navigate through the haystack and find the info it needs?
I have noticed that Jev is not so good at chess nor being a Decision-Model-as-a-Judge. It sucks at chess, and it does not perform well enough to replace GPT 6.1 Sol as a judge in the domain I am working for
But switch is not only for Jev! It is for any decision model including the ones supported by the pi-ai package
So once you can generate enough data with a better model, you should be able to pick an open source one and fine-tune it on your use case to perform better! And the model should be trained in a harness like switch's to be especially performant and cost-efficient
I wanted to get this out because I find the idea very exciting, at least for benchmarking. Jev is currently not there, but decision models obviously will be there at some point, and we can take them there! There are also a lot of potential gains from picking a specific domain, and creating a looping agent that does a job for 100x cheaper!
Let me know if you have any ideas, I am quite keen to try this on new domains!
Repo: github.com/osolmaz/switch
Robert Zapp retweeted
I've been claudemaxxing a bit too hard lately. I needed more compute, more accounts, more guardrails, and now I just need more sleep.
This video is a chaotic spiral into hill climbing. Consider it the start of a series. Only gets weirder from here.
Robert Zapp retweeted
There is an absolutely WILD story in this @GergelyOrosz podcast episode
1. An unnamed CTO of a cash-flow-negative company was debating doing layoffs. AI
2. He says "I don't want to hurt morale like that, I won't do it"
3. A few months later, he says to @GergelyOrosz that he's cash flow positive. WTF. How?
4. The company had been licensing a component from a licensing company that was critical for their business.
5. One day, a dev thinks "what if we got an agent to build our own?"
6. The agent does so well that they produce a BETTER, FASTER version of the component they were licensing.
7. They debate open-sourcing it, and they ask the licensing company about it
8. The licensing company panics, and says "we will pay you NOT to open-source this"
9. They also say "we'll also pay you extra so we can re-license it"
10. So a big outlay became a recurring windfall, thanks to the ingenuity of a dev asking the right question.
Individual contributors have never been more powerful. Think bigger.
youtu.be/mQu0zN4IvSI?si=P_v9…
Robert Zapp retweeted
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...]
Robert Zapp retweeted
The problem isn’t that AI is intelligent. It isn’t. The problem is that too many adults (but not all) are at once mechanistic, imitative, unimaginative, stupid, formulaic, influenceable, & gullible, arrogant beyond their skill set, & too expensive to maintain.
Robert Zapp retweeted
Halfway home, you remember the checkout bug. Your computer is still on at the office.
AgentsRoom's Mobile-Desktop Sync puts your AI coding agents on your phone: type the fix, the agent does it on your computer.
More details: agentsroom.dev/features/mobi…
#ClaudeCode #AIAgents
Robert Zapp retweeted
It's OVER for accountants? Look at these numbers.
Mercor put 12 licensed CPAs, averaging 5.5 years of experience, up against frontier AI on four accounting tasks. The humans averaged 37%. Claude Opus 5 scored 100% across all 20 attempts, finishing each in under 10 minutes. Most human attempts took 30–180 minutes.
Just 18 months ago, the best AI couldn't beat the average accountant on these tasks. Now it beats even the best participant. Mercor wanted to measure how much AI could help accountants. But AI alone scored perfectly, leaving no room for improvement.
Robert Zapp retweeted
@dhh
Now Omarchy has a photoshop level photo editing tool. Thank you @robbietilton for the awesome Compositor!
It has a theme sync specifically for Omarchy Linux. But the standard functions will work on Fedora, Ubuntu as well.
Fully open-sourced on:
github.com/ZacharyZhang-NY/O…
Robert Zapp retweeted
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
