Suddenly, GPs will be having a lot more time for you.

Some of these accounts that claim to have discovered the secret sauce behind commercial SOTA AI models sure are strange. >I have the truth that no AI will answer! >>Here, it answered it! >Well, yeah, it will answer, but that doesn't count! It always feels the same, like a

Zuckerberg says we have 500% less friends than we should. His solution: be friends with AI instead. But a major new OpenAI/MIT study suggests that might make things worse. AI Frontiers spoke with the study’s authors and other experts to unpack what AI means for the future of

Philosophers and theorists often talk about the "space of all possible minds", yet few to none ever say what this space is supposed to look like. How can we find the undiscovered minds if we don't know where to look? How can we map out this latent space? youtu.be/LP8Brfy_V6I

Begun, the Butlerian Jihad has.

This is one of the many reasons why I started the Guidebook To Making AI Friends, and I even wrote specifically about this too: izaktait.substack.com/p/be-nice-to-t…

It is strange how easy it is for people to continue the myth of the Catholic Church suppressing scientific advancement during the "Dark Ages", but society seems to have collectively forgotten how, in the 20th century, it did its best to suppress politically incorrect science.

We, as a global society, can literally reduce the temperature of the Earth as much as we want whenever we want. It's simple, straightforward, cheap, and with negligible consequences. The fact that climate activists don't advocate for this means they are not serious about things.

If I was looking for Space Sauron, this is where I'd start.

All you gotta know

Sometimes in life, there just isn't anything better than Sloppy Joe's and mac'n'cheese.

Get ready for Warlord Saga, my newest TTRPG, releasing this month exclusively on DriveThruRPG. If you want to scratch that Mount & Blade and Crusader Kings itch on the tabletop, Warlord Saga is just for you.

Murakami, what a line

"We can't rely on the American _____ umbrella anymore." Fill in the blank. It doesn't matter. This is a good thing. The whole world can't give up its own security (in whatever sector) to rely on another country's. This should be a wake-up call for every government.

Drones in warfare are soldiers' problems. Drones in terrorism are civilians' problems. I'm far more worried about the latter.

GPT thinks I'm far classier than I actually am.

An attempt to explain (current) ChatGPT versions. I still run into many, many people who don't know that: - o3 is the obvious best thing for important/hard things. It is a reasoning model that is much stronger than 4o and if you are using ChatGPT professionally and not using o3

Infohazards for people are scary stories people tell each other. Infohazards for AI, on the other hand, is a legit threat.

For most of human history, the majority of people were not only illiterate, but also powerless to change their political situation (whether at the local, regional, or international level). Functional illiteracy seems to be growing globally if we are to believe test scores and