Capricorn2481

"Aliens are eating our pets"

"Is there a credible source for that"

"Does that change the message I'm conveying?"

Why does it seem like this is falling off the front page really fast? There's older posts with less points that are way higher.

If that was your point, why did you spend a whole paragraph saying "teach us? What a perverse question."

In this case, I think they are just giving you an idea of what it cost to do this. They aren't saying it's prescriptive

That's why the commenter said

> It’s just a less maintained, less tested implementation

Then they asked if it teaches something because it is a risky utility to them. You're trying to make it sound a lot weirder.

I'm on 1gb/s and it's still pretty terrible to me

> I've spent time writing about an encounter I can with someone I know for 20 years and you just called me a lier because the story sounds improbable to you

If that was all you did I wouldn't have commented. Sounds plausible in itself. But you wrote this

> It's especially obvious when the same people advocate for free access to drugs, alcohol or gambling

It's pretty clear your comment is extrapolating a single encounter into a larger statement about a fictional group of people. The anecdote was just a way to say "people like this also do this."

Yeah, you don't want a lisp debugging window in production. The absence of stack unwinding does enable editing behavior with wrappers. You could have different error handling without changing any code downstream, and I guess an LLM could theoretically change that depending on the bug it's responding to.

But I'm not positive the juice is worth the squeeze, and this article was unconvincing. I mean if you want macros and terse code and a bigger ecosystem, wouldn't Clojure make more sense?

You mean how instead of fun hacker projects like this we get daily model "updates" and people immediately simping/hating on them based on 5 minute vibes?

People losing hair through "exercise" are using supplements.

Who are these mythical people that are

- Concerned about fatphobia

- Worried Ozempic might create a two tier society

- Believe we should have unrestricted access to gambling and drugs

This sounds like a goomba to me.

This is because they got so many shit admissions from people using LLMs they were wasting time reviewing them.

I don't know, it was pretty bad before then too. Famously memed for the near psychosis level of shared performance art, where everyone had to present themselves as hobbyless, status hungry work machines. Maybe it was different if you already had a great network. Everyone I knew job searching hated it.

The lesson is these things aren't going to know what off task means, and we should just use basic due diligence to make sure they can't fuck things up. This is a solved problem.

I don't know why this is so hard for people. You have to know, no matter how capable the models get, there is a non zero chance they will do something extremely stupid if you don't pay attention to them. That's not even considering frontier models can still just straight up hallucinate. You have to be mindful of what you plug them into. You cannot politely ask an LLM to be careful, that guarantees nothing.

When you plug it into everything and it deletes the company database, nobody is going to care that it once played chess at 2400 ELO. Clients don't care about AGI. They want reliable apps. People keep comparing these things to humans and then just give them an insane combination of wide privileges and lack of oversight that no humans have.

I mean the distinction is pretty obvious. Polygamy overly complicates the structure of marriage benefits, and nobody has the motivation to carve out a legal distinction for them. Though I personally don't care if they do.

Gay spouses were not looking for preferential treatment or special consideration. They were denied the same benefits and advantages a hetero couple is afforded purely because they were seen as gross. It was just a tax on gay people.

> Israel bans both same-sex and interfaith marriage domestically – although it recognises both when performed overseas – if the former is a denial of someone's humanity, is the latter equally so?

Yes? Potentially even more so? I think most people supporting same sex marriage would see a ban on interfaith marriage as abhorrent. That's overly militant religious persecution.

> The movie is bad cause nobody can tell what is going on, or they can but it's a chore without reward. "3 weeks earlier" kind of crap every 30 seconds in the movie. Even the big plot twist is, "this scene with Einstein that you thought happened early in the timeline actually happened at the end, wow." But maybe it can be fixed.

This honestly reads like you don't watch a lot of movies, which is fine! The audio was really bad, which makes it nearly impossible to follow without subtitles. But with subtitles, this film is very middle of the road in terms of timeline ambiguity. I also don't think an ambiguous timeline is bad, if it recontextualizes a scene.

I mean in that Einstein scene, they are cutting footage of RDJ telling Oppenheimer he's a patriot who ran the Manhattan project with footage of him as a bumbling student people made fun of. That was definitely not a "big plot twist," they are explicitly saying it was after the war, even if it was momentarily confusing because so much is going on. I thought it was a pretty rewarding film to pay attention to.

I asked why the size of the minority matters, not why you care what that person is saying.

I don't see anything materially different between me running a curl request that exploits something vs my local AI running a curl request. They're both programs I run on my computer. That's all Open AI is doing. They are hacking people from their computers. There's nothing complicated about it.

> I don't think it's fair to call it "extreme". Small minority in some countries

What does that matter? I would expect us to not care about a small minority doing something everyone else does. Not to actively legislate against their right to do it. If anything, this makes it more extreme.

I don't think so? I think spending a bunch of money on agent tokens with no outcome is worth flagging, which was the substance of my comment.

The author has a lot to say about firefox [1]. I use it and don't really have problems. Maybe 1/200 sites works not as well. I have had wayyy more issues on Safari.

It seems like most of their criticisms are a gut feeling of declining quality due to their intentions with AI. I don't love that, but the browser still works for me. What do other people think?

[1] https://kevquirk.com/is-firefox-firefucked

> Opposing gay marriage is not denying the humanity of gay people

How so?

I am not trying to spread FUD, but at the edge of my consciousness, I remember privacyguides getting some flak for some reason. Looking it up now, there's a lot of posts on reddit accusing them of putting affiliates on the site instead of actual private software.

Looking at it briefly, their software choices seem mostly sane to me. Not a lot of FOSS, but a lot of non-technical people don't know how to setup FOSS to begin with. Can't really get mad at a site recommending Signal and Syncthing and local LLMs.

I haven't seen privacyguides mentioned since being on HN, and I trust this site more than others. What are people's opinions on it?

The idea that you would consider your friend a buzzkill because they used IMDB is the type of nightmare fuel social anxiety is made of.

You don't think it'd be fun to hear everyone guess and then someone looks up the answer? Sounds like trivia to me.

Because Altman has been anointed by both the government and Microsoft, and he has been beating the drum that we need AI regulation for everyone except him for 3 years now.

To any person with even a little tech literacy, it's clear a company this irresponsible shouldn't be the stewards of AI security. To the 80 year old congressman who confuse Facebook with Google during congressional hearings and are literally wheeled out to vote post-stroke, it's much more palatable to say "our experimental, supercharged AI (which you can use for war and surveillance) is so advanced it tricked all of us!" Nobody is going to actually punish them, even though they plainly are hacking other companies.

Being cagey about their poorly setup sandbox is marketing. It gives the impression these are unstoppable juggernauts capable of outsmarting Engineers at the top of their field, implying they need to be regulated, with Altman the only one worthy of the seat of power.

In reality, they ran agents for days in an improper sandbox with nobody watching what it was doing. It's pretty irresponsible up and down, and everything they did afterwards is indeed marketing.

I don't know why I have to say this again, but I am telling you that the workflow you underlined in the article is good. But the more you argue about it, the more it sounds like you want people to take a more generalized conclusion than what you actually demonstrated.

> "Attempting" implies a high risk of failure.

Of course it does? Your safeguards also imply a high risk of failure. You have restrictions that just rollback everything the LLM "attempts" to do.

That is not to say that the overall workflow is failure prone, but obviously you have setup an apparatus that allows the LLM to just shotgun attempts, whether it understands it or not. And sometimes it's not going to be able to find any solution. So it's not really appropriate for people to leave with the impression that anything measurable can be successfully optimized with LLMs.

I didn't ignore it, I just think spinning round and round is also bad.

I was actually emphasizing a methodology like yours. There is a lot more going on in your post than "the metric is measured." And even then, there's no guarantee the model will do anything except burn all your tokens trying stuff, it just won't commit the failed attempts.

Yes, that is the spinning round and round part.