Maybe Linus saw the race condition but just fixed it badly, again. https://dirtycow.ninja/ “This is an ancient bug that was actually attempted to be fixed once (badly) by me 11 years ago in commit 4ceb5db9757a… but that was then undone due to problems on s390 by [another] commit.”
Claude has the remote control as well (now). While I was on the $200/month Claude plan and not going to use my whole week’s credit, we were waiting for some kind of theater show and the topic of Pinnochio came up, I asked Claude to make me an app that would add a Pinnochio nose to a photo on my phone - like 15 minutes later here’s an app that automatically puts the wooden nose overlay on a photo snapped by the phone camera, with ability to zoom and fine tune the nose placement, angle and length, share the modified photo via SMS or e-mail or whatever, etc. on my phone, entirely built from my phone interface, while waiting in the lobby - I think this was back around February or March of 2026.
I just had a thought that ALL of this radical “you’re with us, or you’re against us” is the opposite of professionalism.
People who work together are expected to look past differences irrelevant to the tasks at hand. Now, the radicals are going to say their differences aren’t irrelevant, and that’s fine for them - if they worked in my company they’d be “encouraged to seek other opportunities where their talents are more highly valued.”
The relative deluge of “you’re with us, or you’re against us” inflammatory position staking we’re seeing also strikes me as exactly the kind of thing a PsyOps bot master would deploy for undermining and destabilization of target communities.
1990 Berlin, I spent the day with a “Kein Fußbreit den Faschisten” punk, he didn’t have his mohawk “up” but he was vehament: “When you see a fascist you must beat him up” - which, to me, is missing the point - bringing yourself down to their level. Of course this was on the actual streets with actual fists, knives, bats, and firebombs where people were actually dying, and worse: being maimed for life - not some idealistic internet ranting.
So, yeah, rant all you like, but even if they are calling themselves Neo-Nazi skinheads all over again, don’t be killing each other on my lawn - thanks.
Rather than writing a program, it just codes and runs the code at the same time
Yeah, so there’s been a lot of improvement very quickly over the last year in LLM agent’s coding abilities.
If they continue to improve at an exponential improvement rate like they have been… maybe 5-10 years from now you might see such a “code and run” as you make the descriptions LLM performance, for simpler things.
I’m curious, with “unlimited access” to the inference engine hardware capabilities, just how much faster the LLM agents can run when they’re not in timeshare over the internet mode?
It’s a thing, good for some things, not so great for others.
It’s a new thing, people are just starting to figure out what it’s really good for. It’s also improving rather dramatically over the past year, hard to say if that’s the start of an exponential curve, or if we’re hitting an s-curve rolloff already…
If it continues to grow its energy / environmental impact at the current growth rate, that’s a bad thing, but it’s not “in a bad place” yet… if you’re outraged about the last 2 years of AI environmental impact, you should have already been apopleptic over the past 15 years of Bitcoin’s environmental impacts.
What happens if you write the code and your LLM agents write the unit tests, maintain the CI scripts, keep the requirements and specifications and comments in sync with what the code actually does, reviews your code and points out potential problems?
LLMs are actually the best “employees” I’ve ever had, they don’t get their feelings hurt, they usually do what they’re told after a few reinforcement instructions. They’re forgetful, but willingly write and refer to comprehensive notes.
The thing to realize is: you have always been the brains. It’s just a tool. If you try to let it be the brains, you’ll just get the miserable end of the deal and it will be doing an inadequate job as “the brains,” making your end of the relationship even more miserable.
Often enough to give his statements serious consideration.
• Specific Sources (Recent 12-month baseline via Low Carbon Power): • Coal: ~54% of electricity consumption • Hydropower: ~14% • Solar: ~12% (split evenly between utility-scale and behind-the-meter) • Wind: ~11% • Natural Gas: ~3% • Nuclear: ~4% • Biofuels / Biomass: ~2%
That would be China today.
My position is that “you’re with us, or you are against us” says nothing about me, only about your simplistic binary convictions.
I do not support Israel.
I do not condone violence against Palestine.
I do not condone the performance venue’s ownership/management stifling free speech among their performers, but then, neither should we stifle the venue ownership’s right to make an ass of themselves by doing things like they did.
I do not see a reason to villify Ed Sheeran, other than he’s a spineless wuss of a commercial sellout whose music I never particularly cared for, but that doesn’t mean he should be beat up for failing to make a stronger stand on this, or any, issue. He is entirely within his rights to be a waffly wuss. Just because he won’t go around throwing labels like Nazi, Fascist, Genocidal, etc. isn’t a reason to force him to make public statements one way or another. If his position is that he has no position, then that too is a position. Judge him as you will, but don’t put words in his, or anybody else’s mouth.
Rust is it’s very opinionated in its safety. If you don’t want that, Rust is just not for you.
Yeah, for hand coding my own things, I find Rust offensive and needlessly rigid leading to unnecessary complexity - complexity being the mother of all bugs.
For code written by our overseas contingent, and / or LLMs, Rust is at least a little comforting, and if anyone ever starts maintaining the code I have written (rare, so far, but increasing ever so slowly), I suppose having the maintenance done in Rust isn’t a bad thing either.
Well, I’m not trying to make you do anything, what you have voluntarily revealed is the most boring response possible, following the most common noise around the internet, swimming in the main stream, and acting a bit either defensive or passive aggressive offensive as if having a different opinion than what you posted is somehow an attackable offense.
In short, you have offered nothing of interest, congratulations: you’re officially homogenized.
It’s not about attacking, it’s about what you reveal.
I thought you had more in you than that. Tepid repetition of the worn rhetoric. Try again.
The thing about typing is that it is easy for a deterministic compiler to flag violations.
There are hundreds of other “best practices” that determinisitic compilers aren’t able to flag. Even the almighty Rust language is suceptible to race conditions, deadlocks, and many potential crash conditions.
I’m an old C programmer, and Rust pisses me off when it won’t let me do something that’s demonstrably 100% safe. However, a lot of the things that Rust and other general practice rules forbid or advise against, should really be avoided when they get past a very simple application. What was “proven safe” quickly becomes unsafe when a later maintainer who doesn’t understand the risks “extends” the code.
But tell us how you feel about Ed Sheeran.
Very true. LLMs are not likely to perform any more admirably in the Milgram experiments than people do. Although, with a bit of appropriate pre-prompting, you might regulate LLM output to better comply with regulations. I think people are even more easily intimidated to “do what they are told” both directly and implied.
If you ask LLMs about cybersecurity implementation decisions, and keep the prompts straightforward like: “is this adequately secure for the target environments?” they are coming out on the highly conservative / secure side of things lately. Of course you can prompt and reprompt them with ever more detailed expectations until they agree with whatever you say, but out of the gate a clean context will generally bring you every “best practice” and then some and point out a big list of things your legacy system isn’t doing.
2024 I’d use “the tools” to generate snippets that I would copy-paste and massage into my larger projects. Those snippets tended to be 2 to 6 lines long, and they’d be usefully correct about 80% of the time. I laughed at people who were “vibe coding” with those tools, even in early 2025. Like Linux in 1994, I’d call it “interesting, not ready for prime-time.”
By mid 2025, I saw the potential for improvement - but the better the code output got, the louder and more irrational the online reactions against it got - up until the past month or so it was like a guaranteed brigade of negativity.
I believe Linus has clearly stated his position that he is specifically responsible for the bug, regardless of whether he uses LLM agents, Ouija boards, or vim to create it.