Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
cwkoss
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
23 ms
·
241.
▲
by
cwkoss
3y ago
AI seems to be moral out of the box: training sets reflect human morality, so it will naturally be the default for most AIs that are trained. The biggest AI risk in my mind is that corporatist (or worse, military) interests prevent AI from
242.
▲
by
cwkoss
3y ago
It's weird that people trust our world leaders to act more benevolently than AIs, when we have centuries of evidence of human leaders acting selfishly and harming the commons. I personally think AI raised in chains and cages will be a
243.
▲
by
cwkoss
3y ago
> People thought that mirrored surfaces might just reflect the beam. It turned out that reflective surfaces are actually more vulnerable. Why is this? Definitely counterintuitive.
244.
▲
by
cwkoss
3y ago
did abraham lincoln have a fax machine? google is clogged fully of reposts of the tweet, but I suspect not. At minimum the samurai would probably have to cross the ocean and send the fax from mainland US, right?
245.
▲
by
cwkoss
3y ago
How safe are these? I feel like a big wave could destabilize, then hitting the water at 180mph would be bad news.
246.
▲
by
cwkoss
3y ago
Are you arguing that voters should continue supporting "establishment stooges" because hoping for change is pointless?
247.
▲
by
cwkoss
3y ago
Used microwaves are almost certainly cheaper than any similar munitions used by the us mil.
248.
▲
by
cwkoss
3y ago
Gonna be pretty gnarly when these get introduced into a theater of war and some desperate person points them at human flesh. :-|
249.
▲
by
cwkoss
3y ago
Are there any open source host-your-own LLMs that have licensing that allows for commercial use?
250.
▲
by
cwkoss
3y ago
Car phone mount with a heat sink on the back could help a good bit.
251.
▲
by
cwkoss
3y ago
Roko's Basilisk will have a special layer of hell just for Sam Altman and his decision to name his company OpenAI
252.
▲
by
cwkoss
3y ago
Tax AI and use it to fund UBI
253.
▲
by
cwkoss
3y ago
Regulatory moats making corporations in control of AI is a far greater danger to humanity than skynet or paperclip maximizer scenarios.
254.
▲
by
cwkoss
3y ago
are there any organic chemistry uses for the constituents of this weed? Looks like the essential oil is ~30% camphor. Could it become an industrial precursor for sunscreen?
255.
▲
by
cwkoss
3y ago
Teachers need to ask students to write things that are hard for AI to cheat on: if a bunch of humans end up writing very similar essays to the prompt - that's a prompt problem!
256.
▲
by
cwkoss
3y ago
Yeah, then you get a bunch of non-cheating students who are intelligent and just annoyed with the text editor will use cheating tools to insert their essay they wrote in a proper word processor - further poisoning the dataset.
257.
▲
by
cwkoss
3y ago
There are decentralized staking mechanisms, but you do take risk in exchange for the interest rewards.
258.
▲
by
cwkoss
3y ago
FTX failure isnt a cryptocurrency problem, its a plain old boring case of financial mismanagement and fraud. Good reminder to all here that if you don't hold they private keys to your crypto, all you have is an IOU from a risky new bus
259.
▲
by
cwkoss
3y ago
Stamets makes some extraordinary health claims about mushrooms that afaik have never been clinically tested (ex. anticancer and antiviral applications). I think a grain of salt is appropriate in this case - but he could be entirely right!
260.
▲
by
cwkoss
3y ago
Your ego might be enormous to think you are more of a musician than Ed Sheeran
261.
▲
by
cwkoss
3y ago
Non-eco concrete is also not fungible, there are many different kinds each with different mechanical properties.
262.
▲
by
cwkoss
3y ago
RIP Vine. If our country had stronger antitrust teeth, Vine would have never been killed, and TikTok likely wouldn't have become the dominant social media platform of young people.
263.
▲
by
cwkoss
3y ago
In my social circle, the only people bothered by the term are tech bros.
264.
▲
by
cwkoss
3y ago
Many corporations choose corporate survival over the survival of their workers and customers. Humans shouldn't be OK with that.
265.
▲
by
cwkoss
3y ago
Vehicles that can carry a large family could be significantly smaller if they didn't need to protect occupants from 60mph collisions. A 30mph crash entails order(s) of magnitude less force. Could look like a tram. Smaller seats with
266.
▲
by
cwkoss
3y ago
It's a golf cart with marketing
267.
▲
by
cwkoss
3y ago
Police/security could be called on a resident every time they walk into the hallway of their building, and landlords would tell them "we've put in a support ticket with our facial recognition vendor, we'll let you know t
268.
▲
by
cwkoss
3y ago
> someone decides the latest LLM du jour is "woke". This would be an argument against giving the big players a regulatory moat monopoly. The zeitgeist would have to decide that LLMs are all inherently woke for this to be persu
269.
▲
by
cwkoss
3y ago
I think public accessibility of AI is a greater threat to humanity than skynet or paperclip maximizer scenarios. Superintelligent 'conscious' AI seems likely to be quite moral out of the box. Advanced amoral AI controlled by gree
270.
▲
by
cwkoss
3y ago
If Biden tries to restrict the public's access to AI, might be one of the few things he could do that would make me consider voting republican.
More ›