pk-protect-ai

5 ms·

31.
by pk-protect-ai 2y ago
According to this very model there a "mere technicalities" differentiate human and AI systems ... Current AI lacks: First-person perspective simulation Continuous self-monitoring (metacognition error <15%) Episodic future think
32.
by pk-protect-ai 2y ago
But for the training it does. You need to communicate gradient changes between GPUs.
33.
by pk-protect-ai 2y ago
Setting humanitarian arguments aside, here it is: Because preventing pandemics is cheaper than ignoring their possibilities.
34.
by pk-protect-ai 2y ago
Yeap, that the thing, they are not ahead anymore. Not since last summer at least. Yes they have probably largest customer base, but their models are not the best for a while already.
35.
by pk-protect-ai 2y ago
Do I get it wrong, that this will be incompatible with OrthoGrad¹ optimizer? [¹] https://arxiv.org/abs/2501.04697
36.
by pk-protect-ai 2y ago
What about NVLink? Does it plays a role here?
37.
by pk-protect-ai 2y ago
Why people tend to nail some stuff into their products? We have been talking about the AI revolution for several years already, and yet there is no IDE or plugin for VS Code that supports multiple OpenAI compatible endpoints. Some, like Cod
38.
by pk-protect-ai 2y ago
Fly.io probably runs it on Kubernetes as well. It can be something in the middle, like RunPod. If you select 8 GPUs, you'll get a complete host for yourself. Though there is a lot of stuff lacking at RunPod too. But Fly.io... First of
39.
by pk-protect-ai 2y ago
Before GPT-3 was public, there was BLOOM 176B, and this model made my skin crawl because it was capable of answering "I do not know." That was an experience of a lifetime. I was honestly impressed and at the same time scared.
40.
by pk-protect-ai 2y ago
I agree here. I have had an experience of scaling WebSockets server to 20M connections on a single server (with this one https://github.com/ITpC/LAppS.git ). However there are several issues with scaling WebSockets, on
41.
by pk-protect-ai 2y ago
why don't you spam the companies who want your data instead? The sites can simply stop gathering your data, then they will not require to ask for consent ...
42.
by pk-protect-ai 2y ago
and every e-mail, and every document in google docs, and every video on youtube ...
43.
by pk-protect-ai 2y ago
Why don't they want to ditch dark matter? It is more plausible candidate to ditch :)
44.
by pk-protect-ai 2y ago
There is probably an E-Paper capable of JS support, however it would be difficult to use for printing due to it's thickness ... https://en.wikipedia.org/wiki/Electronic_paper
45.
by pk-protect-ai 2y ago
also bold take to proclaim, that humans will be required for teleoperation...
46.
by pk-protect-ai 2y ago
Regardless, under current conditions there is no slim chance of rebellion in Russia. Massive conscription, on the other hand, and of course massive losses of the conscripted may cause an effect similar to 1917. However, the next government
47.
by pk-protect-ai 2y ago
To repeat this, there has to be a condition: "the oppressed can't bear it anymore." Economically, they are totally fine. They are just cattle who are fed and expected to behave as the owner requires. Ninety percent of Russian
48.
by pk-protect-ai 2y ago
How do you change the political environment in a dictatorship where ruling class has all the power and majority of the money and controls what you should read, watch or talk about?
49.
by pk-protect-ai 2y ago
Were those really 8k hours? I remember installing some "new MMORPG" over steam (I'm a linux user) and impossibility to even pass tutorial because of the anti-cheat kick-ins at some later point after login. The whole game sess
50.
by pk-protect-ai 2y ago
You can't win this argument, you are using too many big words and lot of text. Dems should lie as reds to win the votes over... Right?
51.
by pk-protect-ai 2y ago
Why is there an assumption that Trump or reds in general will solve this issue? He was a president already, what exactly did he do to fix the situation? The system is built to segregate and separate people into classes efficiently, making t
52.
by pk-protect-ai 2y ago
Wow, 44GB SRAM, not HBM3 or HBM3e, but actual SRAM ...
53.
by pk-protect-ai 2y ago
The variety of devices available on the market surpasses your imagination. Meta's glasses use the same cameras that may be embedded elsewhere in hats and clothing. There are smaller and larger cameras. Meta's glasses simply provid
54.
by pk-protect-ai 2y ago
"nutrition science circles" are the joke on their own.
55.
by pk-protect-ai 2y ago
Meta smart glasses or any camera, the problem is not in what you are using, the problem is how easy it is to dig out the information about people in real-time.
56.
by pk-protect-ai 2y ago
It is really amazing how consistent this model is in demo videos about world object details over time. This spatial comprehension is really spooky and super amazing at the same time. I hope Meta will release this model with open weights and
57.
by pk-protect-ai 2y ago
Thank you for the context. I wouldn't have read the article without it. I mean, it's a pretty good idea for "no runtime," but when I saw the article title, I thought at first "Why????" Honestly, I'm glad I
58.
by pk-protect-ai 2y ago
And now there will be a legal precedent to shut down all CDLs of these "lots of libraries".
59.
by pk-protect-ai 2y ago
lmao: "Your boundary is at hue 168, greener than 85% of the population. For you, turquoise is blue." When there is no choice to select proper color and you only forced to tell if it is green or blue, despite that you see how many
60.
by pk-protect-ai 2y ago
I hope that somewhere around that period of time we will have AI-based "game" engines working at 30-40fps at 4K (of course with upscaling). I mean it might not be game engines per se, but universal, interactive pipelines for audio

More ›