Waymo's fleet hits 4,000 as recalls and edge cases mount

Share
Waymo's fleet hits 4,000 as recalls and edge cases mount

Robotaxis are scaling faster than the edge cases that trip them up — and OpenAI is working overtime to stay on the good side of the state that just slammed the brakes on its data centers.

Waymo has grown its fleet from roughly 700 robotaxis in 2025 to nearly 4,000 today — and the recalls are growing with it. New York Times reporting, carried by Techmeme, catalogs the failures that come with scale: a flooded-roadway recall filed in late April covered 3,791 fifth- and sixth-generation vehicles after an unoccupied robotaxi drove into water in San Antonio, and three weeks later another Waymo entered standing water in Atlanta. During a San Francisco power outage, 1,593 vehicles paused mid-service. The company suspended operations in Dallas, Houston, San Antonio and Atlanta in May over heavy rain and flood handling, and new state rules now require fleets to clear roads within two minutes of an emergency while reporting every system failure, immobilization and hard-braking event.

The uncomfortable math: more vehicles means more exposure to the long tail of conditions no simulator fully captures — and the fixes don't always stick, since the flood recall was declared addressed before Atlanta happened. Waymo has since reopened Dallas to everyone and returned to freeways, so this reads less like a safety crisis than a scaling tax. But with the fleet headed toward five figures, regulators and riders alike will be watching whether the recall rate falls as fast as the vehicle count rises.


OpenAI has written to Texas Governor Greg Abbott pledging its data centers will "pay their own way" — a direct bid to stay in the state's good graces after Abbott ordered a moratorium on new grid connections. The letter, dated August 7 and signed by CTO of Compute Uday Ruddarraju, commits OpenAI to fully funding project-driven infrastructure costs without shifting them onto residential or small-business customers, supporting new power generation with ERCOT, minimizing water use with closed-loop cooling, and disclosing electricity and water consumption to the state. It lands days after Abbott directed ERCOT to verify data center loads before advancing them through interconnection — the sharpest escalation of a nationwide revolt we've tracked all week, with more than 500 US jurisdictions now pausing or banning data centers (we covered the pushback reaching New York and Texas — US data center bans top 500 as New York, Texas push back).

It's a savvy piece of positioning: OpenAI is volunteering for the exact standards Texas is writing, casting itself as the responsible operator the rules were designed for. The commitments are also notably concrete — closed-loop cooling and no cost-shifting are the two demands that have drawn the most community anger. Whether the letter survives contact with actual construction is the test, but as a playbook for labs navigating the data-center backlash, it's the sharpest one we've seen.

What to watch: Abbott's ERCOT verification results in the coming weeks — they'll show whether the moratorium slows real projects or just speculative ones.

As robotaxi fleets scale toward five figures, what's an acceptable recall rate for a driverless car? Tell us in the comments.

Read more

OpenAI's dots go mobile and can start your Codex work

OpenAI's dots go mobile and can start your Codex work

One product update and two benchmark papers this hour, all circling the same question: how much of the work can you actually hand over, and what happens to the part you didn't? OpenAI's dots — the always-on agents that get their own cloud computer — can now be created and configured entirely from the ChatGPT app on iOS and Android, and a dot can start Codex work or pick up an existing Codex thread on your behalf. The October 9 release, billed as the first of a fresh run of updates for dots, mo

Coding agents as a hospital: Cockroach's five-month results

Coding agents as a hospital: Cockroach's five-month results

The agentic-coding debate keeps being framed as speed versus trust. Cockroach Labs just published the most detailed trust-first ledger yet — five months of numbers, including the parts that didn't work. Cockroach Labs ran its codebase like a teaching hospital for five months and says the pipeline merged well over a million lines of code with seven total reverts. In the setup the company calls MOLT Sinai, GitHub issues are patients, merging is discharge, and the humans in charge are Chiefs of M

Today in AI — October 10, 2026

Today in AI — October 10, 2026

A day where the boring machinery took center stage: proof checkers, order books, warehouses and the humble text message — plus fresh arguments about who gets to declare any of it safe. Models & Research * Mathematicians get a reliability primer for the Lean Theorem Prover. A guest post by Thomas Hales on Terence Tao's blog walks through what Lean actually guarantees — and devotes itself to the "Summer of Soundness Bugs," the string of kernel bugs found this summer that let false proofs thr

Malvertising: fake Claude installers ride Bing redirects

Malvertising: fake Claude installers ride Bing redirects

Claude's popularity has made it a lure — and today's campaign shows how much trust an ad can borrow. One story, dissected. Hackers are running a fake Claude download page through Google Ads, and the trick is that the ad's destination looks like a Microsoft domain. Security researchers at Push Security, who dubbed the technique "Adception," found a sponsored Google result targeting people searching for "claude mac" whose click URL was a legitimate Bing search-results redirect — so the ad passe