grumbelbart2

7 ms·

61.
by grumbelbart2 7mo ago
Here is how that looks like: https://godbolt.org/z/rooee4esd
62.
by grumbelbart2 7mo ago
It's also not representative for the whole industry. BMW is profitable with their electric cars, and 18% of their sales are fully electric.
63.
by grumbelbart2 7mo ago
Going all electric with cars would add ~10-15% of electric demand. That's a bit, but not really a deal breaker, and something Norway would easily be able to offset by adding more wind turbines.
64.
by grumbelbart2 7mo ago
Am I not allowed to choose who I buy my stuff from? Or is it only businesses that are allowed to decline services at will. And do you really belive the richest man in the world, owning one of the largest social media sites, is being "s
65.
by grumbelbart2 7mo ago
> What makes Musk different than Blackrock for instance? Open support for right-wing parties and politics all over the world.
66.
by grumbelbart2 7mo ago
> OP doesn't live in the US, so maybe they don't feel this quite as acutely It's the same sentiment in large parts of Europe, and a major reason for dropping Tesla sales. He actively supports right-wing, eurosceptic politi
67.
by grumbelbart2 7mo ago
Your mother's advice sounds terribly selfish, honestly. Our society is pretty much build on the fact that most people are in some way "good" and will not break laws and rules even if they could get away with it. There are ton
68.
by grumbelbart2 8mo ago
We could probably test this. I wonder if the results shift if the question is prefaced with something like "Here is a trick question: ...".
69.
by grumbelbart2 8mo ago
It's true that hire and tenure decisions are under the institution's control. But a lot of funding comes from external sources, and most public funding uses some sort of publication-based metric. There are exceptions, but that&#x2
70.
by grumbelbart2 8mo ago
This goes deeper than the institutions, actually. The KPI for many (non-industrial) researchers is the number of publications and citations. That's what careers and funding depends on. Goodhart's law states "When a measure be
71.
by grumbelbart2 8mo ago
> a could equal b Why not use "restrict" in this case?
72.
by grumbelbart2 8mo ago
This is absolutely what happens. It's even more tricky since our sensory inputs have different latencies which the brain must compile back into something consistent. While doing so it interprets and filters out a lot of unsurprising, e
73.
by grumbelbart2 8mo ago
No, it's to make it easier to dose different kind of biologically active substances. They can have significantly different "recommended weight to eat of this per day", IUs make that sort-of comparable and easier to remember.
74.
by grumbelbart2 8mo ago
Is that even when using the SHA256 hardware extensions? https://en.wikipedia.org/wiki/SHA_instruction_set
75.
by grumbelbart2 9mo ago
Israel is using those methods in their airport security, quite successfully given their threat level. The problem is that it does not scale well and requires very well trained and attentive personnel.
76.
by grumbelbart2 9mo ago
Last time I filled it out it was optional. Nor sure if that changed in the last year.
77.
by grumbelbart2 9mo ago
Why not "just use a different machine for banking" etc. The point is that most people won't do that. Just like with backups, strong passwords, 2FA, hardware tokens etc. Security and safety features must be either strictly enf
78.
by grumbelbart2 9mo ago
Is there a way to compile it to something else than x86, like arm/aarch64?
79.
by grumbelbart2 9mo ago
> (d) AVX512 is quite hardware specific. Some processors execute these instructions much better than others. It's really a collection of extensions, and you get frequency downclocking that's better/worse on different proce
80.
by grumbelbart2 9mo ago
That is an excellent point. The development costs of such low-volume projects are often way higher than the production costs. Having production in house or close often allows tighter, thus faster, prototyping and feedback during development
81.
by grumbelbart2 9mo ago
https://www.youtube.com/watch?v=0c6mcEPhhRg&list=PLMMxav_zz6...
82.
by grumbelbart2 9mo ago
> but my favourite has to be the Rob Inglis recordings Impressive, very nice. Let's see Paul Allen's recording.
83.
by grumbelbart2 9mo ago
Climate change was likely a factor in 2015. https://www.dw.com/en/how-climate-change-paved-the-way-to-wa... > even the repeated blatant lies It is difficult to have a reasonable discourse when starting with such o
84.
by grumbelbart2 9mo ago
Some say it was a factor in Syria as well: https://www.dw.com/en/how-climate-change-paved-the-way-to-wa...
85.
by grumbelbart2 9mo ago
Maybe. But the data is there, imagine financial troubles, someone buys in and uses the data for whatever they want. Much like 23andme. If you want something to stay a secret, you don't send it to that LLM, or you use a zero-retention c
86.
by grumbelbart2 9mo ago
It probably depends on what "The next IBM" means for people. Microsoft is so deeply embedded into companies right now that for larger cooperation it's practically impossible to get rid of them, and their cloud-driven strateg
87.
by grumbelbart2 10mo ago
One of the comments in the article proposes that: Just wrap both counters at 2 capacity (instead of capacity or UINT _MAX).
88.
by grumbelbart2 10mo ago
I found LLLMs to be very good of writing (unit) tests for my code, for example. They just don't get tired iterating over all corner cases. Those tests easily, in LoC, dwarf the actual implementation. Not sure if that would count toward
89.
by grumbelbart2 11mo ago
I've read this multiple times and disagree. If I had an AGI that designs me a safe, small and cheap fusion reactor, of course I would be interested in that. My intelligence is intrinsically limited by my biology. The only way to really
90.
by grumbelbart2 11mo ago
This is a "classic" machine vision task that has traditionally been solved with non-learning algorithms. (That in part enabled the large volume, zero defect productions in electronics we have today.) There are several off-the-shel

More ›