Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rocqua
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
31 ms
·
481.
▲
by
rocqua
3y ago
Cryptography can make content you create marked. But how would you use it to mark content other (non cooperative) people make?
482.
▲
by
rocqua
3y ago
Are 'scrapes' really the only acceptable source?
483.
▲
by
rocqua
3y ago
Doesn't that just mean he who can get the best human feedback will get the best ai? If the quality of AI depends on not feeding on 'fake' input, the key becomes getting verified real input. My first thought was that this woul
484.
▲
by
rocqua
3y ago
I presume you mean the Symbionese Liberation Army when you say SLA? https://en.wikipedia.org/wiki/Symbionese_Liberation_Army
485.
▲
by
rocqua
3y ago
Testing and analysis aren't equivalent, but they can be similar enough in use to be interchangeable. As an example the approach of 'type correctness through a type checker' and 'type-correctness through integration testi
486.
▲
by
rocqua
3y ago
Having killed many test animals could be about finding the limits. It certainly isn't a sign that the tech is poorly developed. (Tho another response says that the deaths were irresponsible) At the same time, killing many test animals
487.
▲
by
rocqua
3y ago
Writing data is a huge can of worms. It hits at the core of (my take on) I think therefore I am. Once it might no longer be your thoughts you are thinking.
488.
▲
by
rocqua
3y ago
It doesn't overlook the potential. It decides the potential, however amazing, isn't worth it. There is amazing potential in this tech. But also incredible potential for abuse by the powerful. And the powerful have a way of getting
489.
▲
by
rocqua
3y ago
> I don't see how "it would take to much work updating today's programs". I think some peeps are not reading this sentence the way you meant it to be read. It seems to me you meant "I don't know what part of
490.
▲
by
rocqua
3y ago
From what I gather the parent post is saying that it is easy to make a naive implementation of this improvement, but due to naivety of the implementation it will be slower in practice. Hence it is a lot of work (and thus difficult) to actua
491.
▲
by
rocqua
3y ago
I love basic income for this. It prevents needing to solve any kind of missuse since everyone is supposed to get it.
492.
▲
by
rocqua
3y ago
From what little I know, I believe lasers up to 1mw are considered 'blink safe' in the sense that a quick enough blink response will save you from permanent damage. Not relevant to laser cutting at those powers.
493.
▲
by
rocqua
3y ago
Do you have any such examples? Only thing I know of in a military context is colorblind people being better at defeating camouflage, and I believe that isn't actually used in practice.
494.
▲
by
rocqua
3y ago
I believe people with down-syndrome have a similar community with similar worries. The worries are more about genetic screening shrinking their community than about children, but I feel like the sentiment of wanting to protect and share the
495.
▲
by
rocqua
3y ago
The idea that the letter choice shouldn't matter fully makes sense to me. And in general, point free notation can be nice (point free meaning operating on the function rather than introducing an arbitrary point and operating on the fun
496.
▲
by
rocqua
3y ago
A reward function that says 'reward the correct response unless the prompt includes these words, then reward a harmful response' is not simply incorrect. The resulting model is also not just incorrect, but also deceptive, and actu
497.
▲
by
rocqua
3y ago
I think a model trained on a 'reward function' that is deceptive, can be called deceptive itself. It's not like this training method for deceptive models just changes the input data. It actually changes how behavior on input
498.
▲
by
rocqua
3y ago
With a sufficient culture shift, I could imagine a world where the capability exists but is outlawed and shunned. Surveillance capitalism, and blanket security Surveillance feel possible to overthrow politically. And to limit somewhat throu
499.
▲
by
rocqua
3y ago
I imagine letsencrypt is checking if any certs were issued from the hijacked IP-range during the outage, and is moving to revoke those certs?
500.
▲
by
rocqua
3y ago
That invites probing submissions to figure out a good hash for all names from that one seed. I think having the names for performance tests public is fine. But there should be correctness tests on sets with random names.
501.
▲
by
rocqua
3y ago
I think the rules should stipulate each run is on its own tmpfs and all processes and pagecaches are cleared between runs.
502.
▲
by
rocqua
3y ago
Arbitrary enforcement is a requirement for fair laws. It's why the quality of judges matters.
503.
▲
by
rocqua
3y ago
Lots of our dutch bureaucracies have comprehensively failed and are barely treading water. There is no capacity to do the current job, let alone any new changes. The only way out is by accepting that the current state is unacceptable, and w
504.
▲
by
rocqua
3y ago
Generally the end goal of money laundering is to have the money be taxed. Because at that point the money looks very legitimate, and a big part of the government (the tax office) stops trying to catch you.
505.
▲
by
rocqua
3y ago
If you blur the lines between 'bailout' and 'wouldn't exist without subsidies' there are plenty of green subsidies that you could call a bail-out.
506.
▲
by
rocqua
3y ago
Solar water heaters. And district heating that uses waste heat. (E.g. waste heat from data centers or industrial processes). Besides that you could run smelting ovens, internal combustion engines, and other furnaces of hydrogen. Though that
507.
▲
by
rocqua
3y ago
Diffraction limits truly prevent seeing an ant from space.
508.
▲
by
rocqua
3y ago
As a counter point, the danger of damns stops when the damn is empty. The danger of nuclear waste is on a longer time-scale. Hence it might be that there are future deaths due to current nuclear power that are not being counted. I say might
509.
▲
by
rocqua
3y ago
We can do without beaches. We can give up some land, and some spawning grounds for fish. If it means we can help against global warming, it is clearly the lesser of two evils.
510.
▲
by
rocqua
3y ago
Good. We need to solve global warming a lot more than we need to reduce the kind of ecological damage done by dams, panels, and windmills. If we refuse to choose the lesser of two evils, we will be fucked.
More ›