Why do you read that first post as an endorsement not an indictment?
13.04.2025 14:35
👍 6
🔁 11
💬 0
📌 0
Why do you read that first post as an endorsement not an indictment?
#vibejam rokosrebellion.com
Elon's majestic Starship, distributed. AI rendering.
I asked Claude to make me a "Starship simulator".
claude.site/artifacts/d4...
(original from xordev.bsky.social)
Which model? Share the session?
Weird, when I ask it recognizes the paradox and doesn't make the plot:
Weird it worked fine for me:
🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣
Officially have my first basic enemy, a snake that is only just visible behind the grass layer and jumps when you try to jump over it.
Think it's time to try and add push-back on injury
Fittingly, the pile of money gets larger when they torch it.
In case you haven't been on Xitter for a while, this is what you're missing.
You can find the preprint here: arxiv.org/pdf/2408.10234
You can find the preprint here: arxiv.org/pdf/2408.10234
You can find the preprint here: arxiv.org/pdf/2408.10234
🕺
🕺
I think the claim was made ~8 years ago: youtu.be/-uyXE7dY5H0?...
when it was clear that nobody thought "we basically have it already".
Applying some charity, this was a talk about a talk he gave a decade ago. I think this point was meant to be aspirational. An argument that X should be possible, so we shouldn't give up on looking for ways to do it.
"If planes could fly like birds then we can teach them to shit on my ex's sandwich"
It is not the "~" doing a lot of work, but rather the "if".
Point here: youtu.be/1yvBqasHLZs?...
The argument is something like "If artificial neurons are as useful for compute as biological neurons, then anything the brain can do in a fraction of a second should be possible with a deep artificial network, if we can figure out how to train it right."
They're sooooo close to getting it...
I game and dev!
bsky.app/starter-pack...
Thanks!!
bsky.app/starter-pack...
bsky.app/starter-pack...
