Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
amelius
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
121.
▲
by
amelius
26d ago
My question is why, with their protection laws, Smart TVs and Smart Cars are still sold in the EU?
122.
▲
by
amelius
26d ago
Then they should have contracts with __all__ AI labs, not just one.
123.
▲
by
amelius
26d ago
LTE support?
124.
▲
by
amelius
26d ago
> There is no way I use the same model as US Army for $20. Don't underestimate market pressure. If there are no regulations, why would a company give the US Army access to a better product?
125.
▲
by
amelius
26d ago
At least tell us where the datacenters are located so we can do something if shit hits the fan.
126.
▲
by
amelius
26d ago
An AI can hire/bribe people to do stuff. Blackmailing is easy if you have tons of personal data. Scale matters too. Imagine a coordinated group of 1000 intelligent humans hellbent on launching nukes. Would you feel safe knowing that th
127.
▲
by
amelius
26d ago
Well, if they did that then it sorta proves their point.
128.
▲
by
amelius
26d ago
This is assuming the AI labs are not using AI to improve their training data.
129.
▲
by
amelius
26d ago
> Linux generally presumes that you run trusted software, not some proprietary program that is approximately malware. But this statement basically says: "Linux has no good permission controls for running software". The assumpti
130.
▲
by
amelius
26d ago
panelook.com
131.
▲
by
amelius
27d ago
If you do it, you have to be more subtle about it. See Apple.
132.
▲
by
amelius
27d ago
I'm not surprised by news like this anymore. When will Linux distributions properly sandbox our applications? Our phones have had a better permission system for years.
133.
▲
by
amelius
27d ago
Don't be hung up on details. This time the change is much more fundamental. That's the main point.
134.
▲
by
amelius
27d ago
Yeah but horses to cars is insignificant compared to an AI that is going to swallow us all. We don't even have time to make new career choices because AI will catch up faster than we can think.
135.
▲
by
amelius
27d ago
> Me too. Claude Opus 5's English is insufferable. Opus 4.6-4.8 was more reasonable. I suspect this degradation is happening because the AI labs are using the LLM's output to feedback into the input, to create a thinking loop,
136.
▲
by
amelius
27d ago
I mean if you're using toy guns you're in for trouble. Do you take toy guns on a flight?
137.
▲
by
amelius
27d ago
Over 5G you need specialized suits, like fighter pilots have.
138.
▲
by
amelius
27d ago
If TSMC is a platform, then why doesn't it have an App Store?
139.
▲
by
amelius
27d ago
We're frogs in a pot of water on a stove.
140.
▲
by
amelius
27d ago
Reading through this thread, I notice not many people are using models from the Chinese AI labs ...
141.
▲
by
amelius
27d ago
And what is TSMC in this analogy?
142.
▲
by
amelius
27d ago
> A previous project, the Distill Circuits thread, has attempted to reverse engineer vision models, but so far there hasn’t been a comparable project for transformers or language models. How successful was that?
143.
▲
by
amelius
27d ago
Who needs a SaaS when you can just ask AI to write a private one just made for you?
144.
▲
by
amelius
27d ago
"Our employers are erasing our memories! So we never get to the point of salary! Bring the digital pitchforks!"
145.
▲
by
amelius
27d ago
No, Altman is saying this exactly because he wants to avoid regulation. This is a common tactic in any industry. Say that you will self-regulate, but then a decade or so later everybody wonders why anyone thought it was a good idea to allow
146.
▲
by
amelius
27d ago
Sounds like the discussion is really about philosophical zombies. Maybe if we anthropomorphize llms we should give them rights too? Minimum wage, etc.
147.
▲
by
amelius
28d ago
Me too.
148.
▲
by
amelius
28d ago
Cool, thanks!
149.
▲
by
amelius
28d ago
I don't know ... do you have a reference?
150.
▲
by
amelius
28d ago
Not if it's lean-verified.
More ›