cimi_

20 ms·

1.
by cimi_ 21d ago
> No, Meta just amplified it to the nth degree. > [...] > Meta just put a big ol' bright spotlight on the worst aspects that *we* were already amplifying, and *perhaps* sped-run that amplification. No, pick one. If it's t
2.
by cimi_ 1mo ago
I like the website, great work! The LLM descriptions are terrible though :) here's the one for Windows 98: > Windows 98 refined the 95 formula, folding the web into the shell with Active Desktop while keeping the familiar look. Its
3.
by cimi_ 1mo ago
I did something similar 10 years ago, but instead of llms I used word2vec to calculate a embeddings of product descriptions and map those to existing categories. The LLM approach is very likely better, but I'm curious what the cost dif
4.
Keyv and friends compromised in active Shai-Hulud supply chain attack(aikido.dev)
251 points by cimi_ 1mo ago|138 comments
5.
by cimi_ 2mo ago
I stopped using Kagi news when I saw it has Russia Today as a source for many articles
6.
by cimi_ 5mo ago
> As an English-as-second-language speaker and writer How do you know it's actually better? I'm not trying to be condescending, but this reads to me like vibes :)
7.
by cimi_ 5mo ago
This is relevant information for this forum, for good or bad Apple has a lot of impact. I find your reaction strange, do you read news to be angry and/or afraid? :)
8.
by cimi_ 6mo ago
It was announced almost a year ago: https://news.ycombinator.com/item?id=43398518 But it was subject to regulatory approval, that's been completed now.
9.
by cimi_ 6mo ago
> They get the team that built it and have more people on the AI initiative who are consumer-centric. Who are comfortable releasing systems with horrible security, while proudly stating they never read the code? And with metrics that can
10.
by cimi_ 7mo ago
I don’t understand how id-ing the buyer helps? What is the age restricted good here? Are you saying that kids now buy their phones with pocket money without their parents knowing? > It's obvious we're moving in a direction wher
11.
by cimi_ 7mo ago
I can't see how anyone can think "the exporters pay the tariff" makes any sense. TBH, we'll never know how many people thought it made sense because it didn't matter.
12.
by cimi_ 7mo ago
Reading the comments here I see almost everyone posting assumes this is a genuine interaction of an autonomous AI with the repo, not a human driving it. IMO this take is naive :)
13.
by cimi_ 7mo ago
I think this is important - these topics get traction because people like to anthropomorphise LLMs and the attention grab is 'hey, look at what they learned to do now'. It's much less sexy if it's not autonomous, if this
14.
by cimi_ 8mo ago
There was discussion about this here a couple of weeks ago: https://news.ycombinator.com/item?id=46458936 Steve Yegge's a famous developer, this is not a joke :) You could say he is an AI maximalist, from your options
15.
by cimi_ 8mo ago
> The historical pattern suggests cognitive abilities shift rather than disappear. Shift to what? This? https://steve-yegge.medium.com/welcome-to-gas-town-4f25ee16d...
16.
by cimi_ 8mo ago
Not specific to this IPv6 event, but I was wondering what happens to public services during these Internet shutdowns? Does everything stop or it's mostly business as usual minus some things? I would imagine hospitals, tax offices etc n
17.
by cimi_ 9mo ago
I read Rob’s message as against the AI industry, triggered by this email - it is ‘AI is bad’. This startup didn’t spend the trillions he’s referencing.
18.
by cimi_ 11mo ago
Lex Friedman recently did an interview with Durov: https://lexfridman.com/pavel-durov/ I listened to bits of it and I was disappointed by the lack of push back from Lex who was supper excited because he got to hang out
19.
by cimi_ 11mo ago
Speaking of agents and tests, here's a fun one I had the other day: while refactoring a large code base I told the agent to do something precise to a specific module, refactor with the new change, then ensure the tests are passing. The
20.
by cimi_ 11mo ago
Your point to not rely on good intentions and have systems in place to ensure quality is good - but your comparison to humans didn't go well with me. Very few humans fill in their task with made up crap then lie about it - I haven&#x27
21.
by cimi_ 11mo ago
Jetbrains IDEs have 'Local History' that keeps a record of all edits to files - I looked in mine now and I have changes from more than 3w ago :)
22.
by cimi_ 1y ago
They probably mean new code not the entire codebase, but even so I think those numbers are ridiculous given my experience. Is there any evidence of this (anywhere, not just MS or Google)?
23.
by cimi_ 1y ago
I will guess that you are generating orders of magnitude more lines of code with your software than people do when building projects with LLMs - if this is true I don't think the analogy holds.
24.
by cimi_ 2y ago
Why does the title say 'in London'? Everything else I've read in the thread refers to Manchester. I agree with other posters, unless there's a clear strategy to cash in on the equity, you can consider it (to put it nicel
25.
by cimi_ 3y ago
Effects of this are felt today - policing of the population is worse in the countries that were either Nazi occupied or communist compared to the UK. My point was that measures like this destroy this 'advantage'. (I don't kno
26.
by cimi_ 3y ago
Labour is worse if you look at the past years. They were pushing for harsher restrictions in the pandemic and they said this bill is 'watered down' by the conservatives.
27.
by cimi_ 3y ago
I am so disappointed by this. I moved to the UK from Eastern Europe and it's hands down better in almost every way if you consider personal freedom, in my opinion. In terms of policing, it's better than most EU states. See this fo
28.
by cimi_ 3y ago
It's not only the tories: 'Labour pledges to toughen ‘weakened and gutted’ online safety bill' https://www.theguardian.com/technology/2023/jan/01/labour-pl... It's very disheartening
29.
by cimi_ 3y ago
I'm curious, who upvotes this and why? If you've upvoted and see this comment, can you please reply? FWIW, I agree with the rest of the comments here, I think the points on the list can easily be gamed and should not be used/
30.
Anomalous tokens: a mysterious failure mode for GPT(lesswrong.com)
18 points by cimi_ 4y ago|1 comments

More ›