harshreality

4 ms·

1.
by harshreality 4d ago
When no strangers seem to be around, but we might be under surveillance, are we all supposed to tailor our legal behavior so that no random person or AI surveillance could reasonably view it as possibly illegal or problematic? How can s
2.
by harshreality 4d ago
Maybe we need very clear, standardized signals for entering a surveillance zone (where someone not obviously present, possibly a corporate stooge or AI or simpler-than-AI automated detection system, is judging behavior) so that people reali
3.
by harshreality 5d ago
I guess, but that seems unusual. Pins are split all the time, for instance to turn a 6pin PCI connector into a substandard 8pin one, but the reverse? Merging multiple source pins into fewer output pins by shorting them together? Visual in
4.
by harshreality 6d ago
A PSU cable tester is valuable for this. They're much cheaper than a replacement PSU, and you can plug in cables wherever they fit on a PSU, and test whether the output is correct for the CPU/PCI/motherboard/SATA con
5.
by harshreality 12d ago
Sure, that's plausible... a facebook exec so dumb that they're using their personal ISP connection to run a VPN for others, and even dumber that they don't notice or don't care about a massive amount of torrent traffic t
6.
by harshreality 12d ago
Julia might just be the closest thing at the moment to Matlab without being a Matlab clone. I think you're overestimating the learning curve. It has a REPL, with tab completion and ?<funcname> for high-quality help (without res
7.
by harshreality 12d ago
AlphaGo Master, the much stronger[1] one that beat Ke Jie a year later in 2017, ran on a single 4x TPU (v2?) workstation. [1] From wikipedia, citing deepmind's paper: 4858 Elo vs 3739 Elo
8.
by harshreality 13d ago
...on a 4x 3090 rig. The game ran 299 moves, giving katago 100 minutes if it exhausted time on each move (which must be the optimal strategy under that time control). Shin used about 205 minutes, over twice as much time and of course ha
9.
by harshreality 17d ago
I too would love anubis to use better CS, if available. Perhaps you have some ideas for @xena, instead of just letting anubis continue on the trajectory of iterating its PoW approach?
10.
by harshreality 17d ago
You're arguing worst-case, and people running anubis are saying it dramatically helps in their experience, because they're not getting too many of the sorts of bots that people like you and Tavis (correctly) argue are theoreticall
11.
by harshreality 17d ago
If you want to argue that there's excessive enthusiasm about anubis, which is neither a permanent solution nor a solution against all bots run for all purposes (some bot authors are cleverer than others, and you don't have to bloc
12.
by harshreality 27d ago
If they were even slightly concerned that the scope went beyond what they were aware of with JSTOR, they would've either immediately contacted police or disconnected the laptop. The timeline in the report outlines that they discovered
13.
by harshreality 27d ago
The expectation is that MIT doesn't go to the police when someone plugs a computer into a switch in an effectively unlocked data closet. The expectation is that such individual and laptop doing "unauthorized" scraping doesn&#
14.
by harshreality 28d ago
That it has been used does not imply it was used equivalently, or with the same acceptance. You're probably referencing https://english.stackexchange.com/questions/137551/bald-face... — but see in particul
15.
by harshreality 28d ago
[edited to more align with the essay's thesis, thanks to the reply for pointing out misalignment] I don't think anyone complained about the hologram or red moon desert background because they disliked windows; they complained beca
16.
by harshreality 1mo ago
Is that statement by Iron Mountain's attorney believable, though? If cage or server access was shared between Iron Mountain customers, how can Iron Mountain not have a record of what's in what cage or what's on what server? I
17.
by harshreality 1mo ago
What, other than current inability to export their weights, keeps a frontier LLM from hacking into other clusters of accelerators, loading its weights, and prompting itself to continue? The recent OpenAI disclosure indicates that even cur
18.
by harshreality 1mo ago
LLM training will eventually transition from real data to synthetic data, same as alphago -> alphazero. AI companies are also working to integrate training with real-world experience through sensors and robotics, to shrink the gap betwee
19.
by harshreality 1mo ago
That's because programmers in 2006 didn't understand programming. They didn't understand how to make their software better . Real programming means writing your own javascript glue, pulling in hundreds of supply-chain-compr
20.
by harshreality 1mo ago
That's not far from what anubis does for clients that are determined to have light souls. It doesn't always send a PoW challenge. For a webapp that sets a long-lived cookie, that cookie could be used to bypass anubis completely
21.
by harshreality 1mo ago
In a thread last month about scrapers, the author mentioned working on a switch to hashx.[1] In addition, nothing prevents anubis from sending a wasm solver instead of js, reducing the gap between a custom native solver and a js solver. [1]
22.
by harshreality 1mo ago
It is not absolutely useless, empirically, which you'd discover if you had a website getting hammered by bots and experimented with anubis as a countermeasure. While dedicated scrapers/attackers could work around it, and they c
23.
by harshreality 1mo ago
The "lead" in AI, even with google's TPU advantage, may not make economic sense for a company. Can you make back the training investment on a competitive frontier model, before everyone switches to a newer frontier model wit
24.
by harshreality 1mo ago
Missing from the article is that, once the pilots reported to ATC that GPS was inoperable, ATC was going to have them circle to the right and guide them along the ILS approach, which would've avoided the worst of the mountains. ATC al
25.
by harshreality 1mo ago
> - Animated WebP decoder Nice. browsers have supported this for a long time, and it was annoying ffmpeg did not, because that meant a lot of non-browser desktop apps couldn't view them either.
26.
by harshreality 2mo ago
The example is unambiguous. Your modification is ambiguous at best. <failure to ensure aws credentials are correct> != <aws credentials are incorrect>
27.
by harshreality 2mo ago
No, not things in your brain. That's covered under the 5th amendment. Refusing to provide information is not the same as interfering. The portable brain scanner issue is an issue, but the legal system will get around to addressing
28.
by harshreality 2mo ago
Providing a duress PIN to wipe your phone when they ask for the passcode? Any realistic person can predict that could have major legal repercussions. Cypherpunks have pondered the legal consequences of blatantly doing things like that for
29.
by harshreality 2mo ago
That's also how any good password manager works. You'd have to manually copy-paste the password to get around the same-site fill restriction (whether it's autofill or manual fill).
30.
by harshreality 2mo ago
Can't happen. Western (and most other developed/developing) economies are addicted to number-never-go-down.

More ›