layla5alive

This may be one of the most out of touch comments I've seen on HN...

Computers haven't changed much besides.. everything..

Just one very small one: pervasive automated multimodal surveillance systems..

Or... HFT

Eventually lawmakers would adapt to obfuscate their changes so that people wouldn't be able to point their fingers directly at them.. e.g. using accounts run by amorphous groups

Being able to git blame politicians and such would be very useful indeed (at first..) :-)

You should be able to skip the OCR step by grabbing the actual text from the windows via OS APIs, unless your PDFs are scanned and not encoding digital text.

This tool has very tiny quants available for each model, and defaults to the tiny quants, so I'm not so sure you... tried it?

:/ The celebration of loss of rigor does seem like revealed preference.

It is clear that at the limit, machines will be more capable in nearly every way. I suppose there might be some room for a few exceptions like working around big electromagnets that machines can't get close to... Does that sound great? Is there something else you had in mind which would be reserved for humans and fulfilling?

If you mean attempts like centralized communism - absolutely - and the failure mode should have been obvious from human nature: centralizing great power incentivizes corruption.

We don't fully know the failure modes of every alternative yet, but that doesn't mean powerful central governments + capitalism is the best alternative.

LLMs have much better cognitive empathy than most humans have limbic+cognitive empathy - the therapists will also be automated - friction like licensing bodies will slow it down but not stop it.

Yes but same should have been true of animals and humans as a whole did not afford dignity or value to animals. Maybe we will get a taste of our own medicine this time.

Because the capitalists have very big guns and a monopoly on violence.

Its very easy to imagine a better world.

Its much harder to imagine the Scrooges in power willfully giving it up so that we can build that better world.

And doom3.exe or quake3.exe for GPU drivers :)

Yes across all devices (mobile, PC, etc.). If it's a performance issue then it's one of those "happy accidents" that is a bug in their favor. Either they don't care about quality or they really like money more than quality, but no reason it can't be both.

> This seems highly unlikely to be a problem. Most of the interesting/dangerous models are too big to fit in a single GPU instance. Once you have to spread across "normal" networking, performance will be crippled. Then there's the problem of billing...

This... just... doesn't matter. There are ways to scale horizontally at the expense of latency.. token/sec may drop dramatically, but then you just make millions of slow instances and in aggregate, you're back in action as a very powerful coordinated swarm...

Scalar Monomania (profit maximization in this case) is the root of so much evil :/

Can you elaborate? I don't think this is a well known topic at all

Just change AI for Corporations and I agree.

ADHD hyperfocus disables most interrupts.. very task efficient, at the expense of all else :/

Hug.

That sounds awful - sometimes pain is unbearable, and moreso because other people are so blind to it.

And the fear of things being worse if I try but fail to end it..

Often times, for me, it feels like surviving more than living let alone thriving.

I miss having good pain free uninterrupted full nights sleep.

What you feel is valid. And going through it in a new place, with no local support, makes it that much harder. I think most people would feel the same pain you feel, if they were in your shoes. I think most people would want someone to listen/care, too. But the burden of sitting with what you're going through is too much/too heavy, so people reflexively protect themselves from really seeing it..

It sucks, and I'm sorry.

Sometimes animals make better companions than humans through thick and thin... fostering comes to mind as a possible productive distraction from the pain, if that's something you might be compatible with. Though you didn't ask for advice (sorry)

Yes, I was saving my 95% usage for the last day. Then they reset it. Now I have lost 95% usage!! Im back to 7 days and got locked in at only 5% usage for the week - that's actually theft.

Your discomfort may be understated - this phenomena seems to be a slippery slope towards a kind of Dunning-Kruger accidental plagiarism via LLM?

Zero VC funded companies in this space will not eventually invade privacy if they're able to. The only way to avoid that is open source and self-hosted - actually owning the equipment AND fully controlling the open software stack. Anything else will eventually be exploited for spying by a VC funded company. Rug pulls always happen - "Don't be evil" always turns into "Be nearly maximally evil" for financial reasons in corporations configured to be exploitable in this way by their shareholders...

What Nvidia B200 Cards?

Much of the 11% who say it isn't are in on the corruption!

My guess as well. I don't like it when humans silently pass off LLM writing as their own.

I also don't like it when humans silently pass off LLM work as their own.

Prompting is not the same as doing the work. There are ways to genuinely collaborate with LLMs, but if someone silently takes full credit for work generated by LLMs, my assumption is they probably vibe coded it.

Ethics come directly from first principles.

This is the first I've heard of these critiques from him, and as someone who loves open source/Linux, and belives it makes the world better in many ways, I think they're pretty valid. The model has major issues in at least two distinct ways:

1. Not paying for the labor corporations exploit. GPL is better here, but still imperfect. MIT/BSD is awful here. It wouldn't matter in an ideal post-scarcity society, but we don't live in that yet, so it does matter. Until post-scarcity, a better model would route revenue from exploitative capitalist uses back to the creators and/or their desired sinks.

2. The open source bait-and-switch - startups using open source to get users and then pulling up the ladder and going closed source and monopolistic once they have achieved scale to enforce/ achieve lock in.

Open claw: aka. open door to a remote privilege escalation potentially granting root access to your computer (and if you're using it "as intended" possibly all of your email/internet logins/accounts, your credit card, etc.) to any text your model ingests from the internet...

It's already true of LLMs in general that they represent a privilege escalation opportunity to any text they ingest. But with human in the loop, and a well-formed sandbox, the blast radius and risk are both reduced..

Convenience is the root of much evil.

California making this law is but one domino, but if you can't or won't see the bigger picture of the cascade of dominos, then yes, you are the naive one. And there is a need to comment with guns ablazin', this is existential for humanity on a long enough time horizon.

This should be the top comment (minus the first line which, while not untrue, probably doesn't help the message)

We need solidarity to push back against this.

The corporate and government systems we have built are the original paperclip maximizers.

They/collectively "we" are creating an enclosure around all of humanity... Systematically eliminating privacy, liberty and freedom.

Increasingly, the matrix has all of us, and it didn't even need to put us in scifi stasis pods - turns out fiat currency + inflation/infinite growth + resource scarcity + digital markets + internet + smartphones/IoT/devices is already a pretty effective matrix substrate..

Wake up, people...