bgwalter

34 ms·

511.
by bgwalter 1y ago
Misanthropic has no issues putting 60% of humans out of work (according to their own fantasies), but they have to care about the welfare of graphics cards. Either working on/with "AI" does rot the mind (which would be substan
512.
by bgwalter 1y ago
That is another waterfall fantasy paired with Applescript/UML delusions. Perhaps they are seeking vibe investors or they'll indeed offer a toy IDE for vibe coders alongside with their real product.
513.
by bgwalter 1y ago
Because CPython and PyPA are dysfunctional organizations in the hands of people who are in the right (often corporate) cliques. Don't expect anything from there.
514.
by bgwalter 1y ago
How can the results not be influenced if Grok for example lists all questions and answers of a particular episode if asked? It is as easy as the lion/goat/cabbage riddle in canonical form.
515.
Is the AI bubble about to burst?(unherd.com)
1 points by bgwalter 1y ago|0 comments
516.
by bgwalter 1y ago
If you are using the local LLM as a sort of Encarta (another software that failed) it might work, as long as you are fine with shallow answers. Proper research that compares many different pages still requires the regular Google search. Goo
517.
by bgwalter 1y ago
Where does the statement that "inference is profitable" come from? Several sources say they are burning cash, I'll quote the most reputable one: https://www.ft.com/content/3c8cf028-e49f-4ac3-8d95-6f6178cf
518.
by bgwalter 1y ago
One is an unambiguous question, the other depends on the tokenizing algorithm.
519.
by bgwalter 1y ago
And the fully interesting and on-topic article has been flagged off the front page.
520.
by bgwalter 1y ago
Fun article: "Known vulnerabilities are showing up at alarming rates because of these tools. It's just getting worse due to vibe coding and AI coding assistants," he said. "If you wanted to know what it was like to hack
521.
by bgwalter 1y ago
LLMs are able to reproduce the entire IP. Sometimes it requires more than one prompt. I've seen examples in the wild where a single prompt was sufficient: https://jskfellows.stanford.edu/theft-is-not-fair-use-474e11...
522.
by bgwalter 1y ago
And the prompt engineers running the agents will be sitting in Bangalore. Or perhaps outsourced to Infosys. Microsoft under Gates at least produced real things. I wonder when Apple gets an Indian CEO to facilitate outsourcing.
523.
by bgwalter 1y ago
He is probably just trying to sound cool after years of being an "AI" apparatchik. GitHub stole our code under the pretense of "democratizing" software development (while people in poor countries cannot afford the plagia
524.
Theft is not fair use(jskfellows.stanford.edu)
62 points by bgwalter 1y ago|130 comments
525.
by bgwalter 1y ago
How can you put it under Creative Commons if LLM output is not copyrightable? The licence contains: "the person associating CC0 with a Work (the "Affirmer"), to the extent that he or she is an owner of Copyright and Related
526.
by bgwalter 1y ago
In other news, radioactivity was embraced to the point that radium was used everywhere ( https://de.wikipedia.org/wiki/Radium_Girls ) and shoe stores were offering x-rays. Today, even the Internet's positive impact
527.
by bgwalter 1y ago
It is not historical: https://kieranhealy.org/blog/archives/2025/08/07/blueberry-h... Perhaps they have a hot fix that special cases HN complaints?
528.
by bgwalter 1y ago
It is worse than that. I'm backed by the thriving $2 trillion anti-LLM industry! Admittedly it is an unfair fight since the pro-LLM people have no backing.
529.
by bgwalter 1y ago
These blog posts are increasingly AI phenomenology with hardly any concrete examples, other than that cute alliteration example that would be easy to write for a human-thesaurus centaur combination. It can even be a paper thesaurus.
530.
by bgwalter 1y ago
Which paper? The quoted part is my own observation.
531.
by bgwalter 1y ago
> several of them are written by well-known developers who’ve been around for over a decade All of whom had greater output in terms of volume and impact before LLMs. And it is always the same people who now get voted to the top as oppose
532.
by bgwalter 1y ago
Coding bootcamps were always a scam. The conclusions this article draws are insane, but probably what the industry wants to hear.
533.
by bgwalter 1y ago
Pianists' results are well known to be proportional to their talent/effort. In open source hardly anyone is even using LLMs and the ones that do have barely any output, In many cases less output than they had before using LLMs.
534.
by bgwalter 1y ago
"AI" could certainly replace Dohmke. It excels at writing such meaningless articles.
535.
by bgwalter 1y ago
The CEO will likely be fine, Trump also announces movements of nuclear submarines on Truth Social. I'd be more concerned about non-public dealings that Trump might have learned from Roy Cohn, but these are probably off limits for disc
536.
by bgwalter 1y ago
President Obama: https://www.politico.com/story/2009/03/gm-ceo-resigns-at-oba... Sen. Warren: https://www.pbs.org/newshour/economy/embroiled-scandal-wells...
537.
by bgwalter 1y ago
Paraphrased: "Confronted with such extreme potential damages, The Pirate Bay might need to adjust its strategy. It would set an alarming precedent." It is amazing how shamelessly these LLM thieves argue.
538.
by bgwalter 1y ago
It does not give us better docs. The claude.md files are a performative action for the employer. No one will write APUE level documentation because everything is rewritten constantly in those code bases anyway and no one either has the skil
539.
by bgwalter 1y ago
They add claude.md files because they are forced by their employers. They could have done that years ago for humans. I also see it in mostly spaghetti code bases, not in great code bases where no one uses "AI".
540.
by bgwalter 1y ago
Liz Reid staked her career on "AI" working in search. Lo and behold, a blog post by her confirms that "AI" is working. I've seen many outrageously wrong summaries that were contradicted sometimes by articles on the

More ›