Professor, software developer, maker, mentor, fire eater. Pronouns: he, him, his.
Maryland, USA
Joined May 2008
- Tweets2.8K
- Following2.4K
- Followers1.6K
- Likes7.2K
Bill Pugh retweeted
basically, the era of "amazingly smart engineer, just hates to talk with users of the software; shield this person from having to do so with a PM/EM" is over at Amazon
And Amazon is always about maximizing ROI with everything, including sw development
I pay attention to them because of this
Bill Pugh retweeted
When @AnthropicAI published a report this week on how AI is making formal software verification practical at scale, it contained an experiment with deep @BrownCSDept roots. Learn more (and read the paper) at Brown CS Blog: blog.cs.brown.edu/2026/10/26…
Bill Pugh retweeted
Required reading for every intro to programming teacher, for CS or non-CS majors. cacm.acm.org/opinion/teachin…
Bill Pugh retweeted
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
Bill Pugh retweeted
Important context from NYT a few days back: “Before now, Ukrainian drones using Starshield flew as fast as 250 miles an hour with a stable connection. The formal limit had already been 100 m.p.h., but it had not been enforced, General Biletsky said.
That leaves open the possibility that the United States could assert that no change had occurred in the system even as Ukraine experienced changes on the front.”
Bill Pugh retweeted
Breaking: Elon Musk ordered the Starlink slowdown that blindsided Ukraine on.wsj.com/4sgcpBv
Bill Pugh retweeted
"Our initial evaluation suggested that GPT-6-Astra's score was in the late eighties followed closely by Opus 5.5. But, as we started to investigate the mistakes that the models might have made, the models, especially Astra, insisted that they were right and that it was the expert-supplied answers that were erroneous. When we went back to the experts it turned out, in multiple instances, that that the experts found some subtle errors buried deep inside their calculations. As they revised their answers, Astra's score started to inch up and it finally ended up with 98% (49/50)."
Bill Pugh retweeted
Mike Flynn resigned as National Security Advisor in Feb 2017 after lying to VP Pence about his pre-inauguration calls with Russia’s ambassador on sanctions.
Here’s Flynn sitting right next to Putin at a 2015 Moscow gala.
Bill Pugh retweeted
Calling it now - has to be a commercial arrangement with whatever registrar runs Slovenia's .si tld
Bill Pugh retweeted
There's been a lot of debate about AI vs human code review recently. Here's some actual data and observations from inside Meta:
- Engineers now push ~3x the code they did a year ago. Human review capacity obviously didn't triple.
- In an analysis involving 535k+ reviewed diffs, the production-incident rate for AI-reviewed diffs was 1/50th that of human reviewed.
- The revert rate for AI reviewed diffs was 1/3rd of the human reviewed.
- Median review wall time dropped by 35% compared to human-reviewed diffs.
- Experts then looked at every production incident caused by AI-approved diffs. A human reviewer wouldn't have caught any of those issues anyway in a manual review.
- Meta's internal coding agent lands ~3,000 diffs a week, and 59% need zero human edits.
Human review as default for most code changes is already obsolete, and might even be riskier.
Source in comment below.
Bill Pugh retweeted
This one is very exciting for me!! Came via my CS100 professor who pioneered syntax directed editing (aka IDEs) before Turbo Pascal and then my own contribution to Visual C++ that led to Visual Studio to VSCode.
Fall 1983 got this manual and an 8" floppy for my first programming class in college. By senior year I was working as a research assistant on a generalized tool for editors in any language and wrote my first paper. I owe a lot to those Teraks.
A lot of people used these freshman year including (maybe) @Recklight @tballmsft @GhemawatSanjay @puneetster and a few others on here!
Bill Pugh retweeted
It is extremely clear at this point in AI development that, regardless of risk or revenue or any of the other stuff discussed on X all the time, things are just going to keep getting weirder.
Just super, super weird.
Bill Pugh retweeted
Barack Obama at his last press briefing with reporters, 2017:
"I have enjoyed working with all of you. That does not, of course, mean that I've enjoyed every story that you have filed, but that's the point of this relationship. You're not supposed to be sycophants. You're supposed to be skeptics. You're supposed to ask me tough questions. You're not supposed to be complimentary, but you're supposed to cast a critical eye on folks who hold enormous power."
Bill Pugh retweeted
There's been a ton of talk about the role of humans in code review, and when and how humans should be signing off on changes.
I believe that, long-term, humans have no role in routinely reviewing code.
Bill Pugh retweeted
I've been crystalizing some advice for first-year college students: "Advice to the New College Student in the AI Era". Thesis: "the purpose of going to college is to emerge as close to middle-aged as possible".
docs.google.com/document/d/e…
Bill Pugh retweeted
What is prove2.me, the formalization platform that enabled Anthropic's formalization of Fermat's Last Theorem (anthropic.com/research/forma…)? I'll describe more below, and also share the story of how it came out of a class that I wasn't supposed to teach... (1/12)
Bill Pugh retweeted
Rethinking Maths education: A perspective from Olympiad and higher education. ~ Baptiste Serraille. proofsandprompts.com/2026/09… #AI4Math
Bill Pugh retweeted
Honoring Tim Curry. ❤️🖤 Paul Soileau, Jake Shears, and Luke Evans perform a special rendition of “I’m Going Home.”
Bill Pugh retweeted
This week I had the honor of speaking to Princeton’s entire incoming undergraduate class to address their AI anxieties. I had three messages for them — good news, bad news, and a note of optimism. Here’s a condensed version.
The good news
We have enough evidence now to conclude that the shrill predictions of rapid, massive job loss were misplaced. Even in a field like software engineering where AI has been rapidly adopted, its effect has been to shift, not replace the role of the human (see the “decide-execute-deliver” framework normaltech.ai/p/why-ai-hasnt…)
Similarly, the panic about what to major in is also misplaced. There will be enduring demand for computer science, philosophy, and just about everything else. (In fact, AI companies hiring philosophers has been a big recent trend.)
The bad news
AI seems to help senior people much more than juniors. I can use AI for coding because I spent 25 years learning how to code, which lets me supervise coding agents effectively. (See my post on the “growth cycle” vs the “dependence spiral” xitter.gq/random_walker/status/2…)
You are in a bind — you can’t offload your skill-building to AI, but you’ll graduate into a market where employers will expect you to get work done with AI. We never faced this dilemma. As a result we haven’t figured out how to revamp our classes to help you do both. You’ll have to help us figure it out. And you’ll need to somehow resist the constant temptation to turn to the shortcut machine.
The hope
My point is not that AI is bad for learning. It’s an incredibly flexible tool. Is the internet good or bad for learning? Depends — are you using it to find research papers or waste time scrolling? I use AI every day for learning. The key is to use it to increase, not decrease your cognitive load. To learn deeper, not faster. There is no learning without the cognitive sweat. I try to make sure I’m mentally exhausted at the end of the day. I do feel that AI lets me push myself harder than I ever could before, and I have a vision that as AI continues to advance it will enable human-AI “co-superintelligence“. (I talked about this at the end of my ICML keynote. normaltech.ai/p/what-will-be…)
There’s a big, under-appreciated reason why people may have very different experiences and opinions about using AI for work — are they using it for tasks they’re already an expert at, or tasks they can’t do themselves? The former leads to a *growth cycle* and the latter leads to a *dependence spiral*.
When I use AI to do something I’m an expert at, like coding, I treat it as a tool. I can build quickly, maintaining an understanding of the code, knowing that if necessary, I can fix the code myself. It feels empowering. It frees up my time to think about the complex, judgment-oriented parts of software engineering that I can’t or won’t delegate to AI. That means my own skills improve rapidly, and I get to climb the ladder of complexity and develop higher-level skills, much more so than when I write the code myself. I feel in control. I can lock in and achieve a flow state — when AI is working, I’m reviewing, building understanding, and planning the next steps. I never get the feeling that the tool is about to replace me. This is the growth cycle.
(Of course, the growth cycle is not automatic. I still need to exercise agency to use AI responsibly. But it’s the same challenge with any productivity-enhancing technology, and those who’ve navigated such transitions before are well-equipped to navigate it with AI as well.)
On the other hand, if I use it for tasks I don’t understand and haven’t learned to perform myself, I have no choice but to treat it as a superintelligence. If something breaks, the best I can do is ask AI to fix it and hope for the best. I generally can’t evaluate the quality of the output myself. The only way to find out if it's any good is if and when the work is ultimately reviewed by an actual expert. The experience is confusing, unsettling and disempowering. And forget about flow state. By over-relying on AI, I risk losing whatever skill I had at the task in the first place, even if it boosts productivity in the short term. This is the dependence spiral.
It’s no wonder that entry-level workers and students preparing to enter the workforce find themselves in a bind. To compete with the AI-enabled productivity of more seasoned workers, they must adopt AI themselves, but doing so risks the dependence spiral. I have some thoughts on solutions that I will share in later posts, but I think having a clear diagnosis of the problem is a useful first step.
