RSSAmplifier

Blog

Chris von Csefalvay

Chris von Csefalvay is an AI researcher and computational epidemiologist specialising in post-training for language models, agentic AI systems, and reinforcement learning from verifiable rewards. Author of The Craft of Post-Training: A Practical Guide for AI Engineers and Developers (No Starch Press).

chrisvoncsefalvay.comRSS feed ↗20 posts

Latest posts

The view from Skeleton Coast

Egocentric video for robot training now exists at a scale that is genuinely hard to hold in your head, and I have spent this week thinking about everything that is not in it. Two things prompted that. Mattie Fairchild asked, more or less, what your actual differentiator is in a market where everybody is selling the same commodity. And we are releasing CERES tomorrow, which is our answer to a…

The right to become more

There is a peculiar moment in writing a book when something that has existed for years as notes, code, arguments, tracked changes and annoying my unusually patient and kind editors becomes an object with mass. A manuscript can always be revised in one’s imagination – a printed book sits on the table, irreversibly typeset, and begins making its own arrangements with the world. About 20lbs of that…

The War for the Soul of AI

When I was perhaps five, I asked my mom where the soul was. She told me it was between the head and the heart. Not being particularly good at metaphor, I took this rather more literally than she intended. I worked out that anatomically, this would place the soul somewhere around the left shoulder and upper chest region. For years afterwards, I took enormous care of that area. I favoured my left…

The post-training instrument cluster – Part III

Tip Hey, I’m writing a book about this! I’m actually writing a book about this stuff. It turns out there isn’t a lot of literature on how to do post-training at the level too big for single-GPU laptop-sized hobby projects and requiring enterprise reliability on one hand, but not quite at the scale of multi-team distributed post-training you’d get in foundation labs. That’s a problem, because a lot…

Five wild guesses for 2026

There’s an old parable about a drunk searching for his keys under a streetlamp. A passerby stops to help, and after some fruitless searching asks, “Are you sure you lost them here?” The drunk replies, “No, I lost them in the alley. But the light’s better here.” I think about this story a lot when I read AI predictions. Last year, I made five predictions for 2025 , and they held up reasonably well.…

The post-training instrument cluster – Part II

Tip Hey, I’m writing a book about this! I’m actually writing a book about this stuff. It turns out there isn’t a lot of literature on how to do post-training at the level too big for single-GPU laptop-sized hobby projects and requiring enterprise reliability on one hand, but not quite at the scale of multi-team distributed post-training you’d get in foundation labs. That’s a problem, because a lot…

The post-training instrument cluster – Part I

Tip Hey, I’m writing a book about this! I’m actually writing a book about this stuff. It turns out there isn’t a lot of literature on how to do post-training at the level too big for single-GPU laptop-sized hobby projects and requiring enterprise reliability on one hand, but not quite at the scale of multi-team distributed post-training you’d get in foundation labs. That’s a problem, because a lot…

Post-training: three disciplines in a trenchcoat

Tip Hey, I’m writing a book about this! I’m actually writing a book about this stuff. It turns out there isn’t a lot of literature on how to do post-training at the level too big for single-GPU laptop-sized hobby projects and requiring enterprise reliability on one hand, but not quite at the scale of multi-team distributed post-training you’d get in foundation labs. That’s a problem, because a lot…

Dorkestration

In the dying days of the Roman Republic, there existed a class of functionaries called nomenclatores . Their job was to whisper the names of approaching citizens into their patron’s ear so that the great man could greet each one as if they were intimates. It was, in essence, human middleware: a layer of intelligence that sat between intention and execution, transforming the vague desire “I should…

The Hype, the Slop and the Craft

Hype, slop, craft: the evolutionary ecology of AI There’s a patch of ground near Chernobyl that botanists call Рудий ліс, the Red Forest. In the immediate aftermath of the 1986 disaster, 1 the radiation killed mostly everything, turning the pine trees a rust-red colour before they died. That’s poetic, but hardly unexpected – not even Polesian pines can withstand a firehose of low enriched uranium…

What I talk about when I talk about agentic AI

Note Originally written on 01 November 2025, this post was published on 09 November 2025 to comply with pre-publication review. The Latin verb agere means “to do, to act, to drive forward”. From it we get not only “agent” but also “action”, “actor”, and curiously enough, “agile”. The Romans understood what we seem to have forgotten: that agency is fundamentally about motion, about transformation,…

Deep kimchi

This is about the time of year when I get the sudden and irresistible urge to make kimchi. Kimchi – and fermentation in general – is one of those things I wonder if we’re somewhat genetically predisposed to figure out as a species, 1 considering most major cultures in the right climates have independently discovered what is a fairly cool optimisation – finding something that converts sugar into…

SOYA - the only benchmark that matters

In 1790, the French Academy of Sciences commissioned a rather ambitious survey. The goal was to measure the distance from the North Pole to the Equator along the meridian passing through Paris, then use that measurement to define a new universal unit of length: the metre. The idea was simple enough: let’s create a standard so objective, so rooted in natural law, that every nation would adopt it.…

Biome engineering

Roughly 541 million years ago, something extraordinary happened in Earth’s oceans. Over a geologically brief period of perhaps 20 million years, the fossil record explodes with an almost obscene diversity of body plans. Trilobites, anomalocarids, the bizarre Opabinia with its five eyes and frontal grasping appendage – suddenly, the relatively monotonous microbial world gave way to an abundance of…

From agents to the Chorus

Around two years ago, almost to the day, I spent an absolutely frantic evening in Berkeley, going through enough coffee to power a mid-sized city, hammering away at my laptop on trying to figure out what comes after LLMs. What to most people was still barely on the horizon at the time has been a subject I have been working on in various capacities for the best part of the past decade, on and off.…

A slow walk out of Dikika Cave

There’s a cave in Ethiopia, in an area called Dikika. At some point, around 3.4 million years ago, an early hominin made some incisions on an animal carcass, leaving some notches on a bone as the makeshift knife cut past the muscle and sinew into the bone, tell-tale kerf marks that speak of the first time one of our ancestors used a tool. 1 What happened in that cave changed everything for our…

Love in the Time of Algorithms

When Pygmalion carved Galatea from ivory, he fell so deeply in love with his creation that he begged Aphrodite to bring her to life. The goddess, moved by his devotion, granted his wish. 1 It’s a beautiful myth about art, obsession and the blurring lines between creation and creator. What the myth doesn’t tell us is whether Galatea charged a monthly subscription fee. 1 Ovid, Metamorphoses , Book…

The end of isotropy and the rise of metadynamic AI

I’ve spent the better part of this weekend putting OpenAI’s latest offerings through their paces - both the newly released open-weight models and GPT-5 itself. Armed with a selection of coding challenges, mathematical problems, and the sort of esoteric research queries that usually separate the wheat from the chaff, I’ve been conducting what amounts to a weekend-long torture test of these systems.…

A little less conversation: why we need to move from prompting to programming

The Greeks loved oracles. The average temple of Apollo, who among others was in charge of soothsaying and predictery, was adorned to the gills with gifts from grateful worshippers whose inscrutable questions got equally inscrutable answers from Apollo’s oracles. None of these were more famous than the Pythia, the young ladies high as a kite on volcanic fumes at Apollo’s temple in Delphi. As one…

Between the motion and the act

There’s a curious terminological schism that nobody talks about at what passes for dinner parties for the AI crowd. When Russell and Norvig crystallised the definition of an ‘agent’ in their 1995 book – “an entity that perceives its environment through sensors and acts upon it through actuators” –, they seemed to have a particular notion of agency that today’s AI agents fit perfectly: they…