ggsp

Very cool tech! Q1 27 is… Ambitious. Good luck with it, we need new interfaces for AI!

Very cool, thank you for sharing that :) I got curious, funny how every single Wikipedia article for languages which use the Latin script is some cognate of “anthology”, including for “outsiders” (excuse the term) such as Finnish, Basque, or Hungarian, while the Dutch (and the Romans obviously) translated it.

Rust in the future? Excuse the flippancy but what is missing nowadays?

I had no clue about anthology’s etymology. Very cool!

It’s The Register, very much their style

My understanding is that they are implying interstellar travel, because of its complexity, would be impossible without some form of consciousness in charge to steer things.

I'd argue that evolution has accomplished plenty of incredibly complex designs without an ounce of consciousness in play—much more complex than anything humanity has made. And I'd add that even we puny humans have been able to create intelligence without consciousness, meaning the concept of one being orthogonal to the other does not at all seem far fetched to me.

I think Blindsight quickly runs into questions which people will disagree on rather strongly, based on their beliefs and world views. Which I'd say is a marker of a good sci-fi book.

Isn't the point of the book that the aliens are not at all conscious?

I know the point you’re making is half facetious—do you feel like LLMs are specifically getting better at planning?

On the Samsung Flip you can definitely see a crease—it's not super annoying but it's definitely there

If you define AGI as "can do the work of a human sitting at a computer, end to end", then I'd say comparing yourself to it on a specific skill is the wrong test. Can you hand it a role and walk away for a day/week/month?

I can’t yet. I think that I'd want at least two things it doesn't have: the ability to retain what it learned yesterday (without me carrying it in the context window and thus micromanaging it), and the ability to prioritize correctly, i.e. tell which of the n things it could do next is the one that actually matters.

Can't say for sure that those are enough, but not having them seems to be most of why I still have to "babysit" these incredible tools.