
Statistician and computational biologist; alum @maehrlab, @cahanlab, @alexisjbattle lab. Views are not those of my employer. He/him. https://xitter.cf/t.co/jaMuFiHjk0
Joined August 2020
- Tweets3.1K
- Following1.6K
- Followers1.4K
- Likes21.6K
Eric Kernfeld retweeted
Okay, so here's the thing.
If you didn't read Bad Blood or the early WSJ reporting about Holmes (which I did, obsessively, because I am obsessed with scammers) you had a completely different experience of this movie.
You *cannot* have engaged with that media, and then
Eric Kernfeld retweeted
This is what I think many have failed to grasp. T2, Epic, Roblox’s competition is going to be OpenAI, xAI, etc.
The big tech companies are already making interactive world models that will compete with game experiences.
This is why Mark Zuckerberg moved his bets away from XR and into AI.
That is the battlefield where cutting edge tech and the network effect will merge.
Games, AI, and tech are really part of the same network effect. Metcalfe’s law states that the value of a network is proportional to the square of the number of its users. In other words, the platforms with the most creators and players generate the most value.
When Wada-san predicted in 2004 that Networks will transform the game industry, he meant it from a technical perspective. But he knew, and I know because I spent a decade with him, that ultimately it was about the Network of players.
And so we return to the start. Games are no longer a content business. Square Enix knows it too, but the sun set on their opportunity to transition to a Network business. There were a few million players of Final Fantasy 16.
There have been 14.45bn lifetime plays of Grow a Garden so far, with 21m+ peak CONCURRENTS (meaning lifetime players substantially higher). Microsoft sees this too, which is why Everything (and thus Nothing) is an Xbox.
The future is unfolding before us, and given that we’re all free to spend our time as we wish (and are apparently spending it watching TikTok and playing Grow a Garden)
Metcalfe’s law is the future the majority of the audience have already decided with their wallets and time that they want for games. Apparently that's growing gardens on Roblox.
Eric Kernfeld retweeted
Enter AI.
There are many misunderstandings here. But we should not misconstrue the Microsoft layoffs with them replacing studios with AI.
What happened with Microsoft was clear: AI is a one-in-a-generation change in the entire digital order. MS must either get on board, or perish.
Microsoft lost mobile. Microsoft thrived in cloud because it threw the kitchen sink at Azure. Microsoft cannot lose at AI. I don't mean making game art. I mean in that it will replace many methods of communication and impact all facets of human life.
Based on Xbox profitability over the last 20 years, it is guaranteed that a dollar spent by Satya on Xbox would either be burned or gained a few cents.
A dollar spent on AI will either net 100x back in the future, or all is lost. Satya made his choice, and it was the rational one.
"[T]wo estimates exceeded the theoretical maximum ... of 100% ... If vaccination reduced dementia [this much], a decline might be expected ... following ... vaccination programmes; however, no corresponding decline was observed in national data from the UK, Australia, or Canada."
We have a commentary out on some really interesting findings by colleagues in UCL and Finland on some of the maths behind some of the high profile shingles -> dementia analyses. In brief +
Paper: pubmed.ncbi.nlm.nih.gov/4245…
Commentary: kwnsfk27.r.eu-west-1.awstrac…
Eric Kernfeld retweeted
If you've noticed a distinct dialect emerging in frontier AI models and struggled to understand what they're talking about, you're not crazy!
We turned to psycholinguistics to investigate what makes the writing style so weird, and why this might be happening
🚨New preprint! 🧵
Eric Kernfeld retweeted
Today @kannanlakshmi and I are launching Tiles to help leaders turn scattered AI wins into company-wide transformation.
We built its precursor at @CockroachDB, which quickly transformed the way the company worked. Within a few weeks, people from every department had created and shared thousands of apps, skills and agents, and made them part of how they worked every day.
Witnessing this rapid transformation at Cockroach Labs gave us the conviction that what’s holding companies back from reaching business outcomes with AI isn’t a newer, smarter model, but an environment where sharing, learning, and collaborating with colleagues is easy. Armed with the lessons from building that first version, we left our jobs — me after 10 years, and Lakshmi after 7 — to build Tiles.
Since June, we've talked to hundreds of enterprise AI leaders. Those leaders have described similar problems: how do you scale the work that your AI-pilled people have produced into the baseline for the whole company?
Tiles is our answer to that question, and it enables teams to turn one person's best agent, app or skill into the team standard: governed, permissioned, on any model and within a company's own cloud.
Tiles is for every employee: the most AI-forward people on your team, as well as the people who just want to get their jobs done. The current conversation about AI makes it easy to believe that the second group is destined to economic failure. We couldn't disagree more. Not everyone needs to keep up with every new model release and harness update just to stay afloat. They just need what their expert colleagues have already built.
AI at work is still in its MS-DOS era. Motivated adopters get incredible value, but how many people are still staring at a blinking text box waiting for someone to help them get started? Or being told to focus on AI by leadership, with no time to figure it out amidst looming, ordinary deadlines? The Windows moment for work AI hasn’t happened yet. But we think it’s coming fast.
We're grateful to be backed by @AmplifyPartners, @Work_Bench, @CockroachDB and a group of experienced angels including @petermattis, @georgebdavis, and @Nate_Stewart.
Sign up for early access on our website, tiles.ai
Hello World!
We're Tiles. We help leaders turn scattered AI wins into company-wide transformation by empowering subject-matter experts to build, share and promote AI agents, apps, and skills that move the needle for your business. Tiles works with any model or inference provider, has governance, permissions and a full audit trail built in, and runs fully managed or in your own cloud.
Our founders built the first version at Cockroach Labs: in 12 weeks, 85% of the company used it weekly and built 2,000+ apps, and thousands of agents and skills.
Now we're making it available for every company.
Sign up for early access today at tiles.ai.
"Autocomplete hits before limit truncated the list."
Claude seems to maximize my confusion over what part of speech each word is. Should I autocomplete hits? Does the autocomplete hit? Are the hits from autocomplete? guhhhhhhhhhhh
Eric Kernfeld retweeted
I cannot let a steam engine analogy go by without bringing up John Wilkinson, perhaps the most underappreciated engineer of the Industrial Revolution.
Watt had figured out how to dramatically improve the steam engine, but he spent years struggling to actually manufacture one. The problem was embarrassingly simple: nobody could make a sufficiently round cylinder.
The casting and machining tolerances were atrocious. Watt complained that one 18-inch cylinder was out of round by as much as THREE-EIGHTHS OF AN INCH. Steam leaked past the piston. He tried packing the gaps with cork, oiled rags, paper, and apparently even old hats. Nothing worked particularly well.
Enter John Wilkinson, a British ironmaster whose immediate concern was manufacturing better cannons.
At the time, iron cannons were commonly cast around a core, introducing defects and misalignment that could make the barrels inaccurate or dangerously weak. Wilkinson developed a process for boring cannons from solid castings, and in 1774 patented a vastly improved boring machine.
The key innovation was almost comically mechanical: a massive, rigid boring bar, supported at both ends, that could cut a genuinely straight, circular hole instead of following the irregularities of the casting.
He adapted this technology to Watt's cylinders. By 1776, Matthew Boulton was boasting that Wilkinson had bored a FIFTY-INCH cylinder accurate to within the thickness of an old shilling.
Suddenly, the steam engine was a commercially viable machine.
And the story gets better. Wilkinson became one of Watt's first customers, buying an engine to drive the blast machinery at his own ironworks. The steam engine was now helping manufacture the iron that made more steam engines possible.
I find this whole episode absolutely extraordinary. The Industrial Revolution was, for a critical period, bottlenecked on the ability to make a sufficiently round hole.
We remember Watt, and rightly so. We remember the invention of the steam engine and the subsequent development of thermodynamics. But the man who made the machinery that made the whole thing possible is practically a footnote.
I think we consistently underestimate how much of scientific progress is actually downstream of advances in tooling, measurement, and manufacturing precision. It's not just that technology precedes theory. Sometimes the decisive breakthrough is that someone finally figures out how to build the thing the theorists have been talking about for twenty years.
Anyway, I think John Wilkinson deserves considerably more credit for modern civilization than he gets.
a few thoughts:
1. pretraining science is kind of like how people built quite good steam engines before they understood thermodynamics. “Science owes more to the steam engine than the steam engine owes to science.”-(L.J. Henderson). if nick is right, which i am pretty sure he is, progress points in the direction of more interpretability, and pretraining was "slouching towards monosemanticity". my guess would be that major breakthroughs in interp also advance capabilities
2. nick is a good friend of mine, safe to say my origin story at openai wouldn't have happened if i hadn't met him, and it's fascinating watching him operate and seeing the early drafts evolve into this. he's the most profoundly happy guy I know- not driven by guilt, activism, or religious terror, but by fascination. he of course understands the stakes of understanding neural networks, but it doesn't seem like the core reason he does it: he just loves the neural nets. afaict he's posting this from some beautiful hideaway in tokyo. there's a lesson in there for all the +EV maxers. i feel very grateful to have met him
3. i do think it's crazy how few people across the world were staring at neural nets growing: most have moved to a different level of behaviorist analysis, thinking about reward functions and test scores and whatnot. it does not bode that well for the major model companies, oai included, that fundamental breakthroughs like this are coming from outside. we've become too much of a pressure cooker focusing on local incentives, studying narrow tracks under old paradigms, mostly doing engineering, rather than fundamental research, and implicitly hoping the next model will figure it out
Eric Kernfeld retweeted
You should be fascinated hitting the power button on your computer.
Every boot on an x86 machine iterates through the entire sequence of computing evolution in a few milliseconds:
Stage 1:
16-bit Real Mode, acts like an 8086 from ~1978. 500ns (ish); about 15-30 instructions. Goal of this stage is to create a writeable stack+heap.
Stage 2:
32-bit Protected Mode, acts like a 386 from ~1985. Tens of millions of instructions; takes forever on DDR5 /w memory training. Temporarily uses L2/L3 “cache-as-ram” to kickstart main ram (Intel calls this NEM).
Stage 3:
64-bit Long Mode, acts like AMD64 from ~2003. Hundreds of millions of instructions; enumerate pcie devices, gpu, bootloader, etc.
Stage 4+:
Secondary core wakeup. Every *single* core has to go through the same process, but by this point it’s faster because Main RAM is warmed up.
Stage 2 is just so neat man. Quite literally the same concept as an APU/cart start. Use the smaller motor (L2/L3) to kickstart the main turbine (Main Ram) and execute the training process.
Is that not just the coolest thing? Next time I’m annoyed waiting at the DDR5 training process, I’m going to imagine it’s spinning up SR71 turbines.
Eric Kernfeld retweeted
10/ To test it, we used dual-guide Perturb-seq to activate all 1,225 pairs of 50 TFs, one pair per cell. Primed pairs, especially those priming each other (bidirectional), more often had “neomorphic” effects, switching on genes neither TF activates alone, as in FOXA3 + MYRF.
Eric Kernfeld retweeted
I've been prototyping an automated peer review project. With current models you get human level peer review for about $0.10 of tokens. For $5 you can rederive every result, check every statistical inference, and audit every paper cited in both directions.
There are about 100m papers in the scientific corpus. Most are not important. Many are sloppy. For structural reasons it's essentially impossible to change published work, even to correct a typo.
For $500m today you can thoroughly audit every paper ever written. Every p-hacked result, every plagiarist, every fabrication, every citation club, everything.
I don't have $500m so instead I build a tree, focusing on the 1% of papers that carry 99% of the signal, doing a cheap first pass looking for obvious problems. I can find 90% of the nonsense for a few hundred dollars of tokens.
Any guesses where the biggest surprises will be found?
Eric Kernfeld retweeted
Thrilled to share that our work reconstructing a cell fate map of the mouse embryo is out now in @CellCellPress! doi.org/10.1016/j.cell.2026.…
Eric Kernfeld retweeted
Then take the same cells out of the mouse and most of that falls apart. In contrast amino acid synthetases barely care whether they’re in a tumor or a dish; tRNA dependencies very much do.
The dish-to-tumor shift is shared across lines though, and some tRNAs push it while others hold it back.
Alden Scientific has opened searches for additional software engineering roles. Good team; good mission; good plan, so please consider applying.
aldenscientific.com/careers
This is awesome and I think it would be a very valuable resource even if used with 2010-equivalent AI tech. I expect virtual cell models will add at least a modest amount of ability to compose different signals and learn interactions not present in the training data.
BREAKING: The U.S. government, Google, Meta & Zuckerberg-backed @biohub are launching a $1.8B push to build the data layer for AI biology.
The goal is to create the largest open, AI-ready biological dataset ever assembled, giving models the data needed to better predict how biology works & ultimately help treat disease.
The coalition is massive:
› Biohub committed $500M
› DOE is investing $500M+ across measurement, modeling & compute
› NIH is bringing $500M+ of federally funded datasets & repositories
› Meta, Google DeepMind & @IsomorphicLabs are jointly investing another $300M
This is effectively an infrastructure buildout for AI biology.
Instead of just building bigger models, they’re attacking one of the biggest bottlenecks in biology AI today, generating standardized, high-quality biological data at massive scale & making it openly available to researchers.
$1.8B committed to turning biology into something AI can increasingly model, predict & engineer.
There must be at least 10,000 humours these days, no?
Eric Kernfeld retweeted
A few years ago, @EliWeinstein6 and @jura_bio wrote the output of a DNA synthesizer as a probability distribution – a family of generative models that you can sample from in real life, Variational Synthesis (VS). But fitting VS models is brutal. Now we’ve fixed that! 1/6
You get a very different view of the role of AI in drug discovery depending on whether you listen to genomics people ("it's a biological world model") versus medicinal chemists ("it's a research process orchestrator").
- srikosuri.substack.com/p/men…
- linkedin.com/posts/ash-jogal…
Eric Kernfeld retweeted
While we're doing rational things by executive order, can we spend more than 0.05% of the federal NIH budget on studying the mechanisms of aging? By all means let's spend a Manhattan project on treating degenerative disease but can we please spend a Manhattan project on preventing old age? Can't get Alzheimer's or Parkinson's if your body never ages past 25.

























