Soyweiser

I have a big head (I needed to pick the bigger helmets when I went gokarting) so yes, there is a huge relationship between IQ and brain size. Don't mock my chunky noggin please.

Considering that the idea of the singularity of AGI was the exponential function going straight up, I don't think this persons understands the problem. Lol, LMAO foomed the scorpion.

(Also that is some gross weird eugenics shit).

E: also isn't IQ a number that gets regraded every now and then with an common upper bound of 160? I know the whole post is more intended as vaguely eugenics aspirational but still.

Anyway, time to start the lucrative field of HighIQHuman safety research. What do we do if the eugenics superhumans goals don't align with humanity?

Not read it yet, but I do enjoy that Harris Dr Who'ed Trump.

I have often thought about just nuking most of my reddit posts and have not done so because I'm easily distracted. So I also feel like the latter.

Saying “I no longer consent to being in a simulation”

Time to create a big mindfuck for the Rationalists. First have an active account on EA/LW and actively participate. Post this research as a talking point. Then leave one message 'lol just to be sure: I no longer consent to being in a simulation'. And then never touch that account + any related accounts ever again.

insanely detailed geek pages. These contain the right answer! But not the specific right answer.

I had a windows problem once, so was looking through all that looking for an answer and saw somebody go 'I need to save my data, which is very important, so how do I fix this problem' and then they got a reply with 'follow these steps with step X making sure they would delete their data'. And this was before the whole era of AI slop so I fear how much worse it has gotten now.

"I know not with what GPT-5 will will reply with, but GPT-6 will reply with 'Unfortunately as an language model I Unfortunately as an language model I I I Unfortunately Unfortunately Unfortunately model model model'" Albert GPTstein.

Fair enough, my bad. No idea it had gotten that bad. But still, I wasn't intending to rag on libgen, just his idea that these things (which also includes education) are free already.

You are mistaken in my reasoning, I'm saying a person with a well paid university position (which gives him access to money and the university library, which I assume pays for access to their books/papers and doesn't libgen or equivs them) should understand that a lot of training material is indeed not free. This being in addition to the university paying him for his own research, and him prob being pretty annoyed if he was replaced with an iSandberg bot and now was homeless. (This is in addition to what Yud said).

Turns out making papers about replacing the earth with fruit is something Anders-GPT can do perfectly well on its own.

Somebody got all his books from libgen and it shows.

In my experience, mice don’t really break since optical tracking replaced rollerballs.

Sadly they do, sometimes they stop being able to click, or do double clicks or other weirdness. I have gone through a few mice over the years. Oddly, the newer a mouse the more likely this seems to be the case. Don't throw away old but working peripherals.

This does suck for all the non-LLM ML stuff that has actual usage. (Even if a percentage of that was also snake oil, or had dubious success rates).