johnb231

8 ms·

31.
by johnb231 1y ago
They don't, and no one is stopping you from opening a gender neutral club. > offensive to others Sound like you have issues. It's a club. They meet up to play board games. Maybe you could start a club of people who are offended
32.
by johnb231 1y ago
If Apple released x86 desktop processors with the same performance as the M series of course they would be popular. M3 Ultra https://browser.geekbench.com/v6/cpu/10919322 M4 Pro https://browser.geekbenc
33.
by johnb231 1y ago
I think you are a couple of years out of date. No longer an issue with the current SOTA reasoning models.
34.
by johnb231 1y ago
Already happened. Google search includes an AI generated response. Gemini prompts return Google search results.
35.
by johnb231 1y ago
All of the current models have access to Google and will do a search (or multiple searches), filter and analyze the results, then present a summary of results with links.
36.
by johnb231 1y ago
Apple is faster at single core and can match or exceed the 14900k at multi core in some applications. They are not for gaming. They are much faster than Intel for video editing due to hardware codecs. The Mac Mini / iMac / Apple S
37.
by johnb231 1y ago
Intel's older 13th and 14th gen CPUs still have the fastest single core performance. But they are operating at their physical limit and tend to stop working after a few months. Unreliable trash. https://browser.geekbench.com
38.
by johnb231 1y ago
MacBook Pro is 120 Hz
39.
by johnb231 1y ago
$3000 on overpriced junk. Any more than $1500 is a waste of money. It's really not that complicated. Just buy a mid range PC with a cheap 1080p gaming monitor and play the games.
40.
by johnb231 1y ago
> An LLM cannot do that, it cannot attack your point very well, it doesn't know how to say you're wrong, because it doesn't care anyway. It just completes your sentences, so if you say "now you're wrong, change y
41.
by johnb231 1y ago
That's right. No one is "annoyed". AMD has the best value 16 GB GPUs on the market. There is an 8GB version of the 9060. Perfectly suitable option for the majority of gamers around the world.
42.
by johnb231 1y ago
https://docs.google.com/spreadsheets/d/1at1k7qIo5dgPp6K1aCrY...
43.
by johnb231 1y ago
From the article: "Many new games come and go, and oftentimes nowadays the servers are pulled leaving the games unplayable or crippled. Most notably, this has led to a “stop killing games” campaign in the EU and other countries; where
44.
by johnb231 1y ago
Their success with Doom and Quake was a confluence of things that cannot be replicated today. Carmack's programming talent gave them at least a year head start versus the competition. They introduced a new genre with no competition. Ro
45.
by johnb231 1y ago
Rage was released in 2011. His work at Meta produced highly optimized standalone VR. Whether you think it's successful or not, the tracking accuracy and latency is extremely competitive.
46.
by johnb231 1y ago
He is casually dismissing an entire major branch of RL research. Nothing to do with Carmack or any "cult" figure. This is not a solved problem. https://huggingface.co/learn/deep-rl-course/en/unitbonu
47.
by johnb231 1y ago
Ask Gemini to explain how it was trained https://g.co/gemini/share/f64c3358d9fa
48.
by johnb231 1y ago
Rage was Carmack's last big game at id Software before leaving. Romero is credited on 27 games since he left id Software. https://en.wikipedia.org/wiki/John_Romero#Games
49.
by johnb231 1y ago
There was no substance to your argument. Here's Carmack's response to a serious criticism from an OpenAI researcher https://x.com/ID_AA_Carmack/status/1925973500327591979
50.
by johnb231 1y ago
Carmack replied to that https://x.com/ID_AA_Carmack/status/1925973500327591979
51.
by johnb231 1y ago
https://x.com/ID_AA_Carmack/status/1925973500327591979
52.
by johnb231 1y ago
> More people need to get away from training their models just with words. They started doing that a couple of years ago. The frontier "language" models are natively multimodal, trained on audio, text, video, images. That is al
53.
by johnb231 1y ago
Generalization is a formal concept in machine learning and is measurable.
54.
by johnb231 1y ago
The Wikipedia article on AGI explains it well enough. Researchers at Google have proposed a classification scheme with multiple levels of AGI. There are different opinions in the research community. https://arxiv.org/abs&#x2
55.
by johnb231 1y ago
No it is not a “gotcha” and I don’t understand how you got that impression. Carmack believes AGI systems should be able to learn new tasks in realtime alongside humans in the real world.
56.
by johnb231 1y ago
No, the joystick part is really not the limiting factor. They’ve already done this with a direct software interface. Physical interface is a new challenge. But overall you are missing the point.
57.
by johnb231 1y ago
> (100x matmul+tanh?) Biological neurons are way more complex than that. A single neuron has dentritic trees with subunits doing their own local computations. There are temporal dynamics in the firing sequences. There is so much more com
58.
by johnb231 1y ago
From the notes: "A reality check for people that think full embodied AGI is right around the corner is to ask your dancing humanoid robot to pick up a joystick and learn how to play an obscure video game."
59.
by johnb231 1y ago
False dichotomy. It's not "avoiding the real harder challenge". It's solving a different problem and it is extremely relevant to real world applications. These are actual hard and open problems to be solved.
60.
by johnb231 1y ago
I think you are trivializing the field of RL research. Games are not a solved problem. Doing that efficiently in real-time is even more difficult and is highly relevant to real world applications.

More ›