Epitaque

That benchmark might have some issues. You prompted the models to generate an image of striking a ring against a crucible. Then you (presumably, manually?) scored the images that depicted an anvil higher than the ones striking something resembling a crucible.

Looks neat, but closed source, sadly.

I wouldn't mind doing my best steelman of the open source AI if he responds (seriously, id try).

Also, your comment is a bit presumptuous. I think society has been way too accepting of relying on services behind an online API, and it usually does not benefit the consumer.

I just think it's really dumb that people argue passionately about open weight LLMs without even mentioning the risks.

He's also got several nudges to enter your email and subscribe to his blog. And mentions getting 50+ emails a day that he has to respond to and mentions he uses AI to respond to them. He seems to like blasting people with AI content, and also as this blog post mentions, gets sent a lot of AI content himself. It's kinda weird, why is he doing this lol

I echo this sentiment. One of my favorite periods of my life was college, actually getting to learn about some advanced topics in CS. Then I graduated and got a job and now I struggle so hard to learn new things (despite lecture videos and textbooks and LLMs existing) without a professor grading assignments/giving exams/that you can talk to, or classmates.

I’m thinking about enrolling in an online college just for fun. Though the problem I have is that I think the Venn diagram of colleges that are online, aren’t expensive, have advanced CS/ML courses, have an experienced professor that you get to interact with is pretty much zero. If anyone has suggestions, do let me know.

A bit off topic, but does anyone know why there's only one Bluesky relay, or can direct me to an article? Even if the hardware requirements are high, I'd think someone would've made a second relay by now.

Incredibly cool. I enjoyed playing it for a bit but was not huge on some of the characters or art. It'd be cool if you added a level editor so we can add our own scenarios