Eric Kerr

How AI empowered Paul Conyngham to create a custom mRNA vaccine to cure his dog’s cancer when she had only months to live. The first personalized cancer vaccine designed for a dog:

@samlambert When's neki sharding going to be available?

don't worry they are doing sharding next!

Software engineering in 2026

i fit a hyperbolic curve to five AI metrics and found a date. one metric is actually going vertical and it's not the one you'd expect ​campedersen.com/singularity

Undersea infrastructure is the single most fragile layer of the global economy yet nobody in tech is paying attention. Here are the facts: > Big Tech secretly owns half of the world’s subsea internet > China’s shipbuilding outpaces the U.S. by 232× > Deep sea mining is already the size of Mongolia > Underwater drones just dropped from $200k → $10k > Red Sea attacks have hit 100+ commercial ships The ocean is the next trillion dollar frontier, and @angelesahr just mapped the opportunity in her latest research deck. See replies for deck access.

Usenet! groups.google.com/g/comp.lang.ja…

Round 2 of Aurora is possible tonight from after sunset to about midnight! Here are some photos from last nights show in @SaguaroNPS West in #tucson #arizona #aurora #northernlights

So far in 2025, @TheOceanCleanup has been collecting an average of 3200 kg of trash for every hour of the year.

As a neurosurgeon I care a lot about road safety. By now you’ve probably seen @Waymo’s stunning safety results (like 91% fewer serious crashes). But they didn’t just publish data headlines. They released the raw CSV files and data dictionaries. I did a much deeper analysis. A fascinating story emerges when you analyze how they’re achieving this. This isn’t incremental improvement - it’s categorical. We’re looking at the potential elimination of traffic deaths as a leading cause of mortality. The intersection breakthrough: Waymo has essentially solved intersection crashes, with 95% fewer injury incidents than human drivers in the same locations. That’s transforming the deadliest driving scenario. The national math: If every US vehicle performed like Waymo, we’d prevent 33,000-39,000 deaths annually and save $0.9-1.25 trillion in societal costs. Even partial adoption at 27% would save ~10,000 lives per year. In terms of magnitude, this would be the equivalent of eliminating every pedestrian death nationally in a year. The physics signature: Here’s what fascinates me: 47% of Waymo’s contacts involve less than 1 mph delta-V. They’re not just avoiding crashes; they’re converting unavoidable incidents into gentle bumps. It’s like having physics itself on your side. We’re not talking about marginal safety gains. The data represents a fundamental shift from harm reduction to harm prevention. The methodology matters: I used their dynamic geographic benchmarks (comparing like-for-like road conditions) and verified the findings hold across San Francisco, Phoenix, LA, and Austin. The safety advantage actually increases in more complex urban environments. Link to raw data below…. Notes on my approach: Analysis based on 96 million miles of Waymo Rider-Only (RO) data through June 2025, utilizing Waymo's dynamic geographic benchmarks to compare Waymo Driver performance against human drivers under similar road conditions and operational design domains. The projections for national impact (deaths prevented, societal costs) involve several assumptions. Given Waymo's zero reported fatalities, the direct serious injury reductions were mapped to national fatality statistics using established NHTSA-derived ratios that correlate serious injury crash rates with fatality rates. This extrapolation assumes that Waymo's observed serious injury prevention capability would translate proportionally to fatality prevention. Societal cost savings are estimated by applying average per-fatality and per-injury economic costs (e.g., medical, lost productivity, quality of life) as published by NHTSA, scaling these national averages to the projected number of avoided fatalities and injuries based on Waymo's safety performance. These figures represent the potential annual impact if the Waymo Driver's safety profile were widely integrated into the national fleet. @ethanteicher

this is one of the most remarkable technical blog posts I’ve ever read

Prescient, from Shel Silverstein in 1981

NEWS🚨: Scientists just captured the first direct image of the cosmic web — the mysterious structure that holds the universe together

Thoughts on AI (warning: long thread). I’ve been formulating these ideas for a long time…. not months, decades. But only recently has the acceleration made the immediacy of them relevant to almost everyone. They are also very incomplete. But here goes.

I can’t stop looking at this. It’s beautiful We were in the 30-40s not just last year. 5 months ago… We’ve doubled the SWE benchmark in less than half a year 🤯

First Personalized CRISPR Therapy Saves Infant with Rare Genetic Disorder A baby named KJ has become the first person to receive a customized CRISPR gene-editing treatment, correcting a rare liver disorder directly in his body. Delivered via lipid nanoparticles, the therapy targeted a specific mutation causing toxic ammonia buildup. Early results show significant health improvements. This milestone demonstrates the potential of personalized gene editing to treat rare diseases. It marks a shift towards individualized therapies, offering hope for conditions previously deemed untreatable. Such advancements reflect the positive trajectory of AI-driven medical innovations.

I am the 3rd person in the world to receive the @Neuralink brain implant. 1st with ALS. 1st Nonverbal. I am typing this with my brain. It is my primary communication. Ask me anything! I will answer at least all verified users! Thank you @elonmusk!

Brick removal on downtown Cincinnati building reveals 19th century facade.

@chronark @TheAnirudh @Pruxis @Cloudflare @PlanetScale @DrizzleORM Curious if you've checked out CF Hyperdrive? It only works for Postgres (right now?) but is proposed to handle these issues: TLS reconnections, and has built-in read caching / SWR

in a way any network request is going to be slower than hitting your memory cache our data is very very easy to cache, but it's just not really possible in serverless functions (yes I know about global scope, that's what we do right now and only get a 10% hit rate) there are more factors, I plan to write a blog about it

to whom it may concern here's some real world latency from @Cloudflare workers to @PlanetScale using their http driver TLDR: last 30 days: 106ms P50, 303ms P99 The latency is measured by wrapping @DrizzleORM , like so: const start = performance.now() const res = await db.query.xxx.findFirst(..) const latency = performance.now() - start - The planetscale primary is in us-east-1 and we have 2 replicas in eu-central-1 and ap-south-1 - Traffic is weighted towards the US but fairly global. - This measures reads only - Planetscale reports the actual query latencies at ~3ms P50, 5ms P99, the difference is network latency. Whether or not you find that acceptable is your call to make, I just wanted to contribute some non-lab metrics.

Founder of Renaissance Technologies, Jim Simons.