bayesmaxing @bayesmaxing

math lover, hacker, independent researcher Joined April 2011

Tweets

643
Followers

144
Following

616
Likes

85K

bayesmaxing @bayesmaxing

3 months ago

It's such a bummer that genetic algorithms (and other evolutionary inspired algos) don't work. My deep-in-the-bones intuition says that they should work but they just... don't.

1 0 3 127 0

bayesmaxing @bayesmaxing

4 months ago

I hate that enabling web search tends to lobotomize the models. I wonder why that is?

1 0 3 177 0

Computer ♥ Records @ComputerLove_

4 months ago

Before Photoshop there was Paintbox by Quantel.(1981)

11 81 566 12K 57

Download Image

bayesmaxing @bayesmaxing

4 months ago

Ideology is a supernormal stimulus.

0 0 2 99 0

Forget about MMLU-Pro, SWE-bench, or GPQA. The real benchmark is whether or not models improve on being so often wrong, confident AND resistant to correction. THAT is the high order bit. If we don't fix that, what else are we doin' here?

0 0 2 123 0

░ perfectloop ░ @PERFECTL00P

5 months ago

𝙲𝚘𝚛𝚛𝚒𝚍𝚘𝚛 🌊🌊🌊 🌊🧊🌊 🌊🌊🌊

14 156 872 16K 108

Download Gif

bayesmaxing @bayesmaxing

5 months ago

God, I wish I had an intuition for hyper-dimensional spaces. Their properties are so strange.

0 0 3 102 0

bayesmaxing @bayesmaxing

5 months ago

When you strip everything else away, what's left?

0 0 2 97 0

bayesmaxing @bayesmaxing

6 months ago

Simple idea I'm thinking about deeply today: Things don't know how to be good on their own. Things left to their own devices don't just magically 'know' how to be good. You have to get in there and shape it - across all dimensions.

0 0 1 122 0

FEELS @FEELSxart

7 months ago

feels

22 299 2K 35K 151

Download Gif

bayesmaxing @bayesmaxing

7 months ago

0 0 2 97 0

Download Image

bayesmaxing @bayesmaxing

7 months ago

You may not like it but this is what peak computing looks like.

0 0 1 82 0

Download Image

bayesmaxing @bayesmaxing

7 months ago

There really is something different about gpt-4o. It's the first time ChatGPT has gotten me to belly laugh. I said once that the real Turing Test would be for an LLM to say something so surprising, so funny, so incisive that it made me belly laugh. That was just a few months ago.…

0 0 2 104 0

bayesmaxing @bayesmaxing

7 months ago

Thinking about how even if all AI development was frozen at today's level, it would still take several years (decades?) to absorb and metabolize all the progress that has been made so far. What does that mean for the future?