rocqua

5 ms·

1.
by rocqua 7d ago
It seems very hard for OpenAI to prove that independence. Since they seem unable to exclude the possibility that their model was trained on transcripts by the two mathematicians.
2.
by rocqua 15d ago
It’s extra hard on Linux because the amount of ways you can reconfigure a linux machine to work around boot- or root-kits, the ways you can modify a running process, and hide what is happening from an existing process. Really, linux is mean
3.
by rocqua 18d ago
Trumps short term actions will have long term consequences on the alliance. Specifically on how the alliance views reliance on the US. Trump has shown how unreliable an ally the US can be. Putting that cat bag in the bag doesn’t reverse the
4.
by rocqua 21d ago
The worry is not that Nvidia isn’t open about their models. The worry is that Nvidia is trying to control the way you run those models. Trying to bake CUDA assumptions into model design, and pushing the software ecosystem to be as Nvidia fi
5.
by rocqua 2mo ago
Your looking at this from a legal perspective. Consider a diplomatic one instead. Model training is expensive. Model distillation is cheaper. If you train models, and that lets your adversary distill them for much cheaper, that gives your a
6.
by rocqua 2mo ago
It’s more like eating your own fiber supplement in public, and then shitting your pants and telling everyone about it. Sure, it’s embarrassing. But it shows how potent your product is.
7.
by rocqua 2mo ago
I especially wonder about the quality of N big models in a delegation harness against M smaller models in a delegation harness with N and M tuned to use the same amount of total compute. (So a few agents of big models against a lot of big m
8.
by rocqua 2mo ago
https://travel.state.gov/content/travel/en/us-visas/tourism-... It states regarding a visa: "A visa allows a foreign citizen to travel to a U.S. port-of-entry (generally an airport) and request per
9.
by rocqua 2mo ago
Under the assumption that this framework is somewhat sane and only shares biometric data of those people who actually request an ESTA, I don't see the issue. This data already leaks whenever you travel legally to the US, because they t
10.
by rocqua 2mo ago
That's a valuable extra distinction between Mathematics and Software. In Mathematics there is much more clarity on what question you want to answer. It's much less likely you get an answer to a question, and then realize that the
11.
by rocqua 2mo ago
It seems to me you hooked onto the wrong part of proofs vs software compared to what OP meant. The difference OP cares about isn’t how much one cares about style. Instead the important difference lies in validation. A proof can be validated
12.
by rocqua 3mo ago
Do you want others to have an identity? Say policemen, someone who borrowed something, the postman, people asking for donations, your doctor etc? And does it make sense that at least some people do want your identity before interacting with
13.
by rocqua 3mo ago
A significant risk with ZKPs is that they can make a lot of explicit control rules palatable by making the side effects much less extreme. If a company can come up with a reasonable reason to check something, then a ZKP enables that check w
14.
by rocqua 3mo ago
The attestor gives you a credential once. You can think of it as a dob (in e.g. days since 1900) with a digital signature. Then when you want to access something age gated, you locally generate a proof that says “I have a credential signed
15.
by rocqua 3mo ago
We definitely need identity attestation in some places. A basic example is when authorising tax return payments. Or when opening a line of credit. Also when getting e.g. a concealed carry license. And similar for government agents executing
16.
by rocqua 3mo ago
Sure, blame the people not the system. Our whole education system is setup as a skinner box for grades. People finding tricks for better grades are massively rewarded. Study methods are optimised for recall within a month. And educators are
17.
by rocqua 3mo ago
Someone else suggested the bell character. But you can also just set tmux pane names or color. Which you can also do from your agent harnass.
18.
by rocqua 3mo ago
The issue is that lobbying extends to PAC and Super PAC campaign donations. Which is pretty close to a bribe. Lobbyist giving opinions, isn’t anywhere near corruption. Lobbyist giving campaign support by the $100,000 is a lot closer though.
19.
by rocqua 3mo ago
This isn’t about keeping people from having the power of frontier LLMs. So tricks that let others have it aren’t a defeat of this policy. This is export control, where the US government seeks to leverage the fact that these frontier models
20.
by rocqua 3mo ago
There's a very strong podcast series about the common failure of Epidurals: "The retrievals" season 2, by serial productions. I found it rather eye-opening. The series includes reporting on a conference of the Society for Obs
21.
by rocqua 3mo ago
See this comment in the same thread for sources: https://news.ycombinator.com/item?id=48541563
22.
by rocqua 4mo ago
Not vacuous, but tautological. Which is different, because tautologies can actually be quite directly informative. Whereas vacuous truths tend to be oblique. Also, “Microsoft is lying” is not a logically stronger statement, because they mi
23.
by rocqua 4mo ago
Which licenses allow usage for training? MIT, BSD, etc likely do. But I would expect it gets weird for all the various copyleft licences.
24.
by rocqua 4mo ago
It doesn’t fully guarantee that. But it guarantees you don’t have a huge class of bugs. And it makes concurrency a lot easier to reason about. No system will likely ever guarantee that software does what you expect. That runs into the halti
25.
by rocqua 4mo ago
Microsoft themselves won’t do that. They are already under severe scrutiny internationally for fear of the US using Microsoft as leverage. They don’t want to stoke those fears. Once they do something like this, everyone who has been saying
26.
by rocqua 4mo ago
This is a rather strong analysis. And especially the point on behaviour change once market growth plateaus was new to me. Thanks! I do want to nitpick on “unregulated free markets”. Because it’s almost an oxymoron. At least if one wants to
27.
by rocqua 4mo ago
At least the servers bit seems very related no? I’d love to know more though.
28.
by rocqua 4mo ago
If this removes people’s access to products (software licenses count as products here) someone payed fir once. Then you should only be allowed to do that if you enable people to continue using the product. Releasing the server code should b
29.
by rocqua 4mo ago
It’s not just vaguely bad because of corruption. But its corruption that very directly increased the price of oil payed by everyone else.
30.
by rocqua 4mo ago
This insider trading isn't hedge-funds working hard to get an edge. It's political insiders trading ahead of public statements. They are getting gains not by dint of being incredibly smart, nor from working very hard. Instead its

More ›