dminik
> If it would be possible to walk to get groceries
It is.
> do people choose to drive instead because it's simply easier and faster?
Some. You also have to factor in the cost of gas (quite high) vs distance vs how much you plan to buy and possibly other factors (like discounts,...).
> They then claim that the goal of exercise is simply not worth it any more given driving is faster (and easier)?
I would think the goal is to get groceries, but yeah. People are quite lazy. The possibility of taking a car or a bus is always there.
Translating LLM proofs to human-ese is instead grunt-work. One that is bound to dissapear in a few years anyways.
Unfortunately, the browser is not a Minecraft server. It's a living standard that has to maintain backwards compatibility. The desire path (React, Svelte,...) will always exist. And so will the ridiculous way they tried to stop that.
The original react documentation is fine, but focused almost entirely on class components. What was there on hooks was either outdated or outright harmful.
It's not until 2023, 4 years after react hooks launched, that the new react docs which focused on them were released.
Of course this needs bootstrapping. But, paying a guy on Facebook marketplace (or whatever) to unpack and turn on your robot for 50 bucks doesn't require superintelligence.
The re-render model just doesn't quite fit with the way most JS libraries work. They hand you a class that you have to keep around. This can be solved with useState, but you have to use this weird construction `const [instance] = useState(() => new Whatever())`. Or even worse if the library needs an element to instantiate.
Then, you likely need some event listeners. They ended up not implementing useEvent, so this has to happen in useEffect. God forbid if you need to rebind the listener because a prop changes (possibly every rerender) or the library does something heavy on adding them.
Then, if a prop changes and you need to do something (like recenter a map), useEffect.
Occasionally you might want to reach for useSyncExternalStore but then get hit by the identity issue. The return value must be the same between renders or you end up with infinite rerenders. Many libraries just do `return { foo, bar}`. Not good enough for react.
Basically every JS library integration ends up with a thousand use effects, various useMemos (and useCallbacks) and I've never got (or seen) a satisfying result. There's always sync issues or too many rerenders or...
A good example is worldnews. For a long time I thought it was a pretty good source of, well, news. But at some point I started noticing that certain topics (cough Negativity about Israel cough) which got a decent amount of discussion were deleted as duplicates of posts that came in later and had 0 comments and could not be seen on hot/top/whatever.
Back to your topic, I think this is in large part due to the death of multiple forums and personal websites and the aggregation of everyone to like 4 platforms.
In the past, if you were banned or didn't like one forum, you could find another. Or maybe just post your thoughts on your own website/blog. But now nobody will see it unless it's on reddit, discord or YouTube.
I don't think we can get that back, but maybe a bluesky(-ish) forum/reddit alternative where you can chose who you interact/see posts and comments from would work well. It would take another collosal fuckup to get people to move and even then... Lemmy was alive during the API/third party apps thing and nobody really moved.
The teams at OpenAI know this as well and yet there was no supervision. Thousands of instances of these advanced systems are allowed to run wild with no oversight.
I have my doubts that the HuggingFace hack would happen if a person was reading the thoughts and executed commands as they happened in real time.
That's the negligence.
You're right though that ethics don't matter into it. But as long as we can't train an LLM to stop picking a sledgehammer to remove a tooth, then alignment is not easy and trivial.
Well, unless of course you want to train your own LLM, or do some biochemistry (and increasingly just regular health stuff) or cybersecurity. These capabilities are not made available for plebs like you or I.
I agree, yes. But it's a bit like saying the only way to not die in a car crash is to not drive. If we're in a situation where using LLMs is unavoidable, I would rather make them safer.
> Well there’s your first problem. A line in the prompt is not a safeguard. Even if you could trust the model - and you cannot - there is always the issue of prompt injection. A proper safeguard means actual sandboxing.
You're right. I did not completely sandbox it and air gapped it. But I also wented it to do some actual work.
If I completely sandbox it, but still leave it the ability to compile stuff, it's just going to build it's own (bad) version of the tool. That's obviously not what I want either.
The obvious thing to me would be for the LLM to notice it's limitations, reason through why they might exist and explain to the user that it cannot do it's job without such and such.
But that brings us to my original comment that these things are over-optimized on completing the task by any means necessary.
For instance, I was playing around with Claude a few days ago and it decided that it was missing a tool and it was going to get it one way or another.
First, it tried apt. No sudo, so no install that way. Tried installing via mise, but it didn't have the permissions. Then moved on to grabbing the source from github and building it.
Should I have included a "DO NOT UNDER ANY CIRCUMSTANCES INSTALL ANY TOOLS"? I mean, I had to after that. But how many other things am I missing? And at what point do the safeguards become so long they get consumed by compaction, or just ignored by the model?
I doubt Google is distributing the Android code to third party OEMs that way.
It feels like the models have been very optimized at getting shit done. But not so much at figuring out what the limits should be.
That is still dangerous and it shows that the models ARE misaligned with what their users are wanting/asking them to do.
Judging by public statements, Google is one of the 3 big western AI companies. Surely they should be rolling in cash and working hard towards AGI.
And yet, for whatever reason, they can't help themselves from further restricting user freedoms on Android. Why?
I don't want to be conspiratorial, but surely it's not money, right? It has to be control. Someone high up at Google just seems to resent people having control over their own devices.