Eliezer

13 ms·

1.
by Eliezer 13d ago
> even with Fable, I've been at the point personally where I am reasonably confident that there's essentially nothing that I am better than Fable at despite generally being substantively above average on human benchmarks If I a
2.
by Eliezer 2mo ago
Why on Earth would anyone take anything this man said to be informative about anything, other than what he thinks will currently sound best that minute to the current Zeitgeist that day?
3.
by Eliezer 5mo ago
You can turn that off in settings.
4.
by Eliezer 7mo ago
The safety teams are trivial expenses for them. They fire the safety team because explicit failure makes them look bad, or because the safety team doesn't go along with a party line and gets labeled disloyal.
5.
by Eliezer 7mo ago
Came here to say the same.
6.
by Eliezer 7mo ago
There's intrinsic limits to vanilla transformer stacks. Nobody knows where they are. We don't know how unvanilla Opus 4.6 or GPT 5.3 are. We don't know what's in development or which new ideas will pan out. But it wi
7.
by Eliezer 7mo ago
Every time somebody writes an article like this without any dates and without saying which model they used, my guess is that they've simply failed to internalize the idea that "AI" is a moving target; nor understood that they
8.
by Eliezer 8mo ago
"2024 AI was a horse". People really like to imagine that the last 6 months constitute their true observation of the new eternal state of the future.
9.
by Eliezer 11mo ago
It would not surprise me at all for bb7 to exceed Graham's number. Just a Kirby-Paris hydra or a Goodstein sequence gets you to epsilon zero in the fast-growing hierarchy, where Graham is around omega+2.
10.
by Eliezer 11mo ago
Thanks to OpenAI for voluntarily sharing these important and valuable statistics. I think these ought to be mandatory government statistics, but until they are or it becomes an industry standard, I will not criticize the first company to h
11.
by Eliezer 11mo ago
Privacy.com has always been a working general solution to this, for me. Disposable CC aliases with spending caps.
12.
by Eliezer 1y ago
I have tried to tell my legions of fanatic brainwashed adherents exactly this, and they have refused to listen to me because the wrong way is more fun for them. https://x.com/ESYudkowsky/status/1075854951996256256
13.
by Eliezer 1y ago
My god, how long has it been since you tried to use an AI model?
14.
by Eliezer 1y ago
Doubling the productivity of 20% of workers, in cases where a lower price doesn't increase demand, can shift prices in the whole system as unemployed artists compete with other artists for wages. AI won't take your job, someone
15.
by Eliezer 1y ago
Translators? Graphic artists? The omission of the most obviously impacted professions immediately identifies this as a cooked study, along with talking about LLMs as "chatbots". I wonder who paid for it.
16.
by Eliezer 1y ago
Poor eyesight is evolutionarily recent (not enough sunlight exposure in childhood, rare to find in hunter-gatherer societies). Baldness won't kill you.
17.
by Eliezer 1y ago
And if it isn't already false it will be false in 6 months, or 1.5 years on the outside. AI is a moving target, and the oldest people among you might remember a time in the 1750s when it didn't talk to you about code at all.
18.
by Eliezer 2y ago
If the LLM did anything besides try to explain the Efficient Market Hypothesis in response, it failed.
19.
by Eliezer 2y ago
Hadn't seen that before, and despite being bog-standard Bostrom it's still more of an attempt to hold a theory than I'd seen associated with him before. Note nonoverlap of writing style and theory with the present post.
20.
by Eliezer 2y ago
I wonder who wrote this? Doesn't sound like Altman's voice. I wonder who theorized this? Altman isn't known for having models about AGI. To the actual theorist: Claiming in one paragraph that AI goes as log resources, and
21.
by Eliezer 2y ago
No major AI advancements for 7 months? Guess everyone's jobs are safe for another year, and after that we're all dead?
22.
by Eliezer 2y ago
This seems like decent alignment-positive work on a glance, though I haven't checked full details yet. I probably can't make it happen, but how much would someone need to pay you to make up your time, expense, and risk?
23.
by Eliezer 2y ago
If you do build a superintelligence, you don't have an ASI, the ASI has you.
24.
by Eliezer 2y ago
Superintelligence is not the same branch of the tech tree as drones or surveillance.
25.
by Eliezer 2y ago
Meanwhile, no luck getting it to build something that reverses a GIF. (Also, weirdly, no luck with finding a working GIF reverser online.) (Trying to reverse this: https://www.tumblr.com/necessary-disorder/7650640081
26.
by Eliezer 2y ago
lol, like the government doesn't have 3 more Mersennes they keep secret so they can verify potential First Contact situations
27.
by Eliezer 2y ago
THAT will help. Thank you to the Nobel Prize committee for this choice.
28.
by Eliezer 2y ago
This is not how nonprofits usually work. This is blatant fraud. I cannot think of any other case besides OpenAI of this particular shenanigan being pulled.
29.
by Eliezer 2y ago
We're not there yet, and the future is famously hard to predict; but it sure feels like you drew a straight line and time continued to move along that straight line even if it's not quite at that destination yet.
30.
by Eliezer 2y ago
These weren't Israeli products. If I were Mossad, I'd compromise anything except an Israeli or Jewish-owned product.

More ›