I have better thoughts I can waste time on

Me too. Unfortunately I don't get to pick my intrusive thoughts.

Sometimes, I get to peer into the AI pilled brain, and it is... uh... not good.

But now if I suffer from Imposter syndrome I can remind myself that some other people take advice from clankers.

https://bsky.app/profile/carnage4life.bsky.social/post/3m634z3r7kk2l

Transcript:

The fact I can login every morning and ask an AI to review all my emails and chats from yesterday then given what it knows about my goals and my role it should suggest what I could have done better is amazing.

Bubble or not, AI is huge for personal productivity and overall improvement.

I have no opinion, but I have to note that I keep reading "KeepAssXC ..."

Beff Jezos and friends have produced something other than tweets. Possibly. Maybe.

https://youtu.be/Y28JQzS6TlE

I have no idea what is good web design. I'll just note makes the waving red, white and blue flag in the background makes the white heading text pretty hard to read.

I think it's more like Alien vs Predator.

I'm wondering about the benchmark too. It's way above my level to figure out how it can be gamed. But, buried in the article:

Moreover, ARC-AGI-1 is now saturating – besides o3's new score, the fact is that a large ensemble of low-compute Kaggle solutions can now score 81% on the private eval.

The most expensive o3 version achieved 87.5%

Also, I the image is perfect. I especially like the Joe Kucan-looking general embedded in the star trek tactical station. The Technology of Peace ain’t what it used to be, is it?

Is that a screenshot from Command&Conquer 4?

I thought it was for Bozo

Who knows. The only thing that came to my mind reading that is the joke "statement made by utterly deranged". And then I realized there is no joke.

Yeah, neural network training is notoriously easy to reproduce /s.

Just few things can affect results: source data, data labels, network structure, training parameters, version of training script, versions of libraries, seed for random number generator, hardware, operating system.

Also, deployment is another can of worms.

Also, even if you have open source script, data and labels, there's no guarantee you'll have useful documentation for either of these.

The first paragraph surprised me. I didn't know there were still some true believers left.