GTP

23 ms·

421.
by GTP 2y ago
I'm doing a PhD about pre-silicon fault injection. I was working on a speeidea to improve current fault injection simulators for hardware designs, but I may have to put it on the side for now, since I stalled for a while now.
422.
Xonsh – A Python-powered shell(xon.sh)
171 points by GTP 2y ago|47 comments
423.
by GTP 2y ago
I'm not familiar with Engel's view, but my gut feeling here is that he was complaining about the way profit was made ("sucking out their very life-blood") and not about profit itself. But, even if he somehow saw profit a
424.
GeForce RTX 5070 Ti GPU announcement(blogs.nvidia.com)
2 points by GTP 2y ago|0 comments
425.
by GTP 2y ago
I have a more practical view: there's nothing wrong in making profit, the important thing is that they are also doing some good.
426.
by GTP 2y ago
Thanks for the heads up, I missed this project! However, on their page they write "Project Timeline: 1 May 2024 - 30 April 2025". April isn't far away, anyone knows what's supposed to happen afterwards?
427.
by GTP 2y ago
Build an AI agent for that :D
428.
by GTP 2y ago
Asking if it is a big deal or not is definitely a question ;) Thank you for providing the information I was missing though.
429.
by GTP 2y ago
But maybe this is actually the reason behind the presentation: he sees this vast effort to transition to PQC, and he's wandering whether it is worth it or not. And for what is worth, I'm also a bit skeptic of the maturity of PQ
430.
by GTP 2y ago
Is there any recording of this being presented?
431.
by GTP 2y ago
Where did you find a Kobo Forma so cheap? A quick check on Ebay lists the cheapest at 150€, but maybe I'm looking at the wrong model.
432.
by GTP 2y ago
But Deepseek R1 doesn't use CUDA, so maybe for this specific case, it isn't a big deal?
433.
by GTP 2y ago
Yes, but we don't have enough people doing that to get quality data. Not many people are building this kind of setup, and even less are publishing their results. Additionally, if I just run a test a couple of time and then average the
434.
by GTP 2y ago
>The last few years of LLM progress have shown we can predict human "reasoning" responses to inputs by modeling likely human responses as if LLM generated. Put another way, most responses are not particularly reasoned, but chai
435.
by GTP 2y ago
Deepseek R1 is reportedly able to 1-shot a flappy bird clone in Python, did you try it?
436.
by GTP 2y ago
> In the last two weeks, I've heavily used Claude and DeepSeek Chat. ChatGPT is much more reliable compared to both Which reliability problems did you face? I heard about connection issues due to too much traffic with Deepseek, but
437.
by GTP 2y ago
This somehow reminds me of a certain killer robot from a Black Mirror episode ;)
438.
by GTP 2y ago
Sorry for being lazy, but I just don't have the time right now to read the paper. Is there in the paper or somewhere else a comparison based on benchmarks of S1 vs R1 (the full R1, not quantized or distilled)?
439.
by GTP 2y ago
But in cryptography, you have a security claim that is based on the hardness of a well known mathematical probelm. So you know that breaking a certain scheme is as hard as solving that problem. If someone finds out how to solve such a probl
440.
by GTP 2y ago
Exactly. Above all, my concern is that, at least from the linked article, I don't see why an attacker wouldn't be able to train a neural network himself to then decrypt the data. Hopefully they addressed this point in the paper.
441.
by GTP 2y ago
Yes, but just throwing in the possibility of a future unexpected breakthrough without grouding this in the situation so far, can lead to wishful thinking. The fact that the list has 5 independent points means that you need 5 independent bre
442.
by GTP 2y ago
Seeing it from this perspective, maybe I now see it: the residual radioactivity inside the reactor may be strong enough to mess with the electronics of a robot you're sending inside. Maybe there are challenges in getting enough shiel
443.
by GTP 2y ago
There are a few things I don't get of the fifth point: why do we need to be able to operate nuclear fusion plants remotely? Wouldn't it be not just possible, but maybe even preferable, to have staff on site to manage the plant, li
444.
by GTP 2y ago
Yes, I get that it is an issue of scale, but that's the issue in the end. A commenter rightfully pointed out that there's no point in giving examples of other things that consume the same or more if at the end of the day we don&#x
445.
by GTP 2y ago
Honest question. I remeber some recent news about Microsoft considering to bring back in use an "old" nuclear power plant to cover their (maybe projected?) cloud power consuption due to AI. If AI's power consumption is neglig
446.
by GTP 2y ago
I took it form one of the Stackexchange answers, I'm not sure what they're trying to say either.
447.
by GTP 2y ago
' “I have no legs” means that I have 0 legs (as opposed to 1 leg or 2 legs), while “I have no leg” means that I have not any leg (i.e., I do not have any leg) ' made me laugh :D Edit: To be honest, I don't see the difference
448.
by GTP 2y ago
Yes, I would agree with this principle, with the caveat that there could be always be corner cases that deserve a special treatment.
449.
by GTP 2y ago
This might be a case of conflict of interest though, as saying instead that Ulbricht did ask him to kill someone could worsen the legal position of the alleged hitman as well, depending on the specific circumstances of the case.
450.
by GTP 2y ago
It is legal if you're in good shape and therefore the risk of that happening is minimal. It is illegal to drive a car under an altered state that makes it more likely to happen. It is a balance between the benefits of permitting someth

More ›