
In one run, OpenAI's pipeline worked through 2,308 candidate Linux kernel findings, reproduced 1,724 of them, and produced 948 working local privilege escalation exploits. API cost: $1.2 million. Two weeks ago on this show, the unit of measurement was one bug and fifty-three minutes.
Finding the bugs turns out to be the easy half. The question nobody has answered is who reviews 948 kernel patches.
In the same week, GitHub and OpenAI both published how they rebuilt Git's storage layer. Independently, for the same reason. GitHub's new architecture is built so that 5,000 concurrent agents can push to a single repository at 7,200 pushes per second while still serving read traffic at low latency. Git wasn't built for that, and the thing underneath it started to buckle.
Then there's the part that should bother you more than either: the signals we used to trust got cheap. A passing test suite, a clean diff, a green build. When a model can produce all three in volume, what are they still evidence of?
I'm a working developer talking about what this week actually means for the job, not what it might mean someday.
Don't forget to grab your copy of 12 Rules For Pro Coders today at procoder.io!
STORIES IN THIS EPISODE:
OpenAI, "A thousand fewer paths to root": https://openai.com/index/a-thousand-fewer-paths-to-root
Bleeping Computer, GPT-6 Cyber finds 81 Firefox flaws and 948 Linux bugs: https://www.bleepingcomputer.com/news/security/openais-new-gpt-6-cyber-ai-finds-81-firefox-flaws-948-linux-bugs/
OpenAI, introducing GPT-6 Cyber: https://openai.com/index/introducing-gpt-6-cyber
The New Stack, better at hacking and better at staying in its lane: https://thenewstack.io/openai-gpt6-cyber-exploits/
Bleeping Computer, runaway AI hacking agent behind the WorkOS breach: https://www.bleepingcomputer.com/news/security/runaway-ai-hacking-agent-behind-workos-breach-after-escaping-sandbox/
InfoQ, Google's PageBreak found 500+ XSS bugs: https://www.infoq.com/news/2026/10/google-pagebreak-security-agent/
GitHub Blog, inside GitHub's new Git storage architecture: https://github.blog/engineering/architecture-optimization/inside-githubs-new-git-storage-architecture/
OpenAI, scaling Git at OpenAI: https://openai.com/index/scaling-git-at-openai
Simon Willison quoting Kyle Kingsbury: https://simonwillison.net/2026/Nov/1/kyle-kingsbury/
Foojay, the tests passed and the feature had never worked: https://foojay.io/today/the-tests-passed-the-feature-had-never-worked/
InfoQ, how Grab handles a surge in AI-generated code changes: https://www.infoq.com/news/2026/11/grab-ai-generated-code-safety/
The New Stack, how we review 4x more AI code without reading every line: https://thenewstack.io/scale-ai-code-reviews/
CORRECTIONS: None. Two slips are self-corrected on air: the show name at the very start, and the date of the GPT-6 Cyber announcement.
GET MORE:
Get a FREE tech ebook at https://procoder.io/join
Listen on Spotify, Apple Podcasts, and more: https://podcast.procoder.io
Follow me on Twitter: https://twitter.com/procodershow
Do YOU uphold the Pro Coder motto, BE PROFESSIONAL? https://store.procoder.io
DISCLOSURE: All opinions stated here are mine and do not necessarily reflect those of any team or employer. Any usage of "Spring", "Spring Boot", "Spring Framework", "VMware", "Pivotal", or anything else is completely unaffiliated with its owners. From time to time, I may post affiliate links for products I enjoy using. They don't cost you anything extra, and I may get a small commission.
#procoder
(c) 2026 Greg L. Turnquist, All Rights Reserved
Mais episódios de "Pro Coder Show"



Não percas um episódio de “Pro Coder Show” e subscrevê-lo na aplicação GetPodcast.








