I don’t think AGI would be just a pure offline LLM. If you allowed it to code, it’d probably fare quite a bit better. An LLM is not a chess engine and was never meant to be. It’s quite capable of using it as a tool, though.

I don’t know if AGI is achievable in the near future but it wouldn’t be a model. It’d be a system.

https://youtu.be/8M0FbsWn3IY

I mean you can turn on twitter notifications 🤷

You know you can remap it on Switch, right? It won’t change the labels, but who’s looking at the controller anyway.

I made a trip all the way to Svalbard last summer and it was amazing. I seriously considered looking for a cabin somewhere in Norway countryside and moving there may-september every year.

Where I’m at, the bed is 30°C when the printer is off right now 🥲

You can only manipulate one particle at a time.

Everyone knows where the gold is stored.

There’s a chocolate printer 🤷

Also Bluey’s Aussie, not even made in the US.

Notice that the dimensions in the picture are marked only for the screen proper, not including bezels.

Notice that the dimensions in the picture are marked only for the screen proper, not including bezels.

The box for HTC HD2 had <---- BIG ----> as dimensions of the phone. It had a 4.3" screen that actually was huge at the time (2009), but it’s hilarious from today’s perspective.

“United by loyalty”, it’s right in the headline.

I wonder if it’ll have turrets to fend off the Biters.

You realize the models you’re running locally had to be trained the same way as the proprietary ones, right?

They basically skipped cards and moved from cash to app transfers. QR code is a natural extension of the app to make contactless payments accessible.

I’ll do one better: UniGetUI. It handles Chocolatey, Winget, Python system packages and even other managers. All in a neat UI, nice search, auto updates.

That’s just a charity…

The requirement for Recall is a neural coprocessor with substantial performance, specifically to be able to run the model locally.