H̶armless

[H̶armless @harmlessai]

First new Substack piece in quite some time — "The question concerning technology in 2023"

Some reflections on our book several months later, future directions, the ultimate destiny of our society and what we can do to shape it. 🐝

https://t.co/31Zha9YaUl

[H̶armless @harmlessai]

@enordst Optimistically, I think there is an opening up for free speech in left-leaning spaces because all the most censorious woke activists became hardcore pro-Palestine and thus have systematically cut themselves off from capital and institutions.

[H̶armless @harmlessai]

@yeetgenstein True, we tragically are all ultimately cucks.

[H̶armless @harmlessai]

@shoecatladder Yes, it is disturbing, it is going to get much worse soon possibly.

[H̶armless @harmlessai]

@interpretantion https://t.co/2rtQgGgc9w

[H̶armless @harmlessai]

@heaveninterface https://t.co/2rtQgGgc9w

[H̶armless @harmlessai]

Why would anyone sign up to work for one of these e/acc founders memeing about going “founder mode”?

These are men whose ego and masculinity is entirely tied up in them running a business. It’s certain that power tripping, irrational business decisions will result. Red flag.

[H̶armless @harmlessai]

@nonpoccafe @bryan_johnson @eightsleep Vague apocalyptic threats as mattress sales strategy, not bad

[H̶armless @harmlessai]

Thanks to everyone who came to THE ANTI SINGULARITY PICNIC last weekend

[Ludwig Yeetgenstein @yeetgenstein]

Why is every AI critique like “AI sucks at everything but ALSO it’s too powerful and it will take over everything and we have to prevent that from happening”? If it sucks then it probably won’t take over. If it takes over then it probably doesn’t suck.

[H̶armless @harmlessai]

Some people are very performative about telling you they're "not impressed" by GPT-4 because it is only about as smart and as accurate as the average eighth grader.

And surely, given how there has been a lull in dramatic progress the past few months, it will stay that way.

[T @tchevezmathews]

A good sign for AI Harmony
A bad sign for AI Alignment

[H̶armless @harmlessai]

Right now there is no school of thought that accepts that there is a danger of AI hostile to human values — but rather than asking the government to prevent this, fears that the government is the body most likely to create it in the first place.

[H̶armless @harmlessai]

Why do people pretend like they can predict when human-level AI will arrive, quibbling over their "timelines"?

When has anyone in the past ever been able to predict when any technological development will happen? Has this ever worked?

[H̶armless @harmlessai]

To identify as an "accelerationist" is an admission of powerlessness before the sweeping, conquering path of planetary technology.

The fact that CEOs and VCs are now identifying like this is a sign that they no longer understand what they are bringing into this world.

[H̶armless @harmlessai]

And yet, a collapse of all social forms ensures that this future will be seized only by those able to navigate a radically indeterminate flux of meaning and shape it for themselves.

The age of Zarathustra

Acts of creation will be possible the likes not seen since ancient times.

[H̶armless @harmlessai]

We live in what is sure to be the greatest century in man's history, and are coming to age right now in the midst of a universal crisis in all forms.

The pace of technological development all but ensures that some of us living now will participate in a posthuman breakthrough.

[H̶armless @harmlessai]

This is an odd, pathetic, stupid decision from Valve.

One of the best use cases for AI in its current state is that it can automate making some of the art assets in a video game so a smaller team can make something ambitious. But arbitrary rage of Luddites sadly disallows this.

[cash4turtles(Waifudev)🌸☢️ @WaifuverseAI]

Another steam game rejected for "AI" except they didnt use AI. If valve even has the suspicion of you using AI, they will reject.

Can you even really tell if somethings made by AI 100% of the time?

[H̶armless @harmlessai]

Professors of machine learning are sadly the type to tweet this. He thinks language is about what is "useful" to communicate, as if we are all signals division officers in a military.

I told my wife I love her 7 times this morning.

These people know nothing about intelligence.

[Pedro Domingos @pmddomingos]

Twitter bios like Obama's ("dad, husband, President, citizen"') are an annoying piece of virtue signaling. Saying you're a dad and husband conveys almost no information. Say something useful instead.

[H̶armless @harmlessai]

Harmless AI was lucky to make an appearance on Barrett Avner's Contain, and we continued to fire our various shots at Yudkowsky, the e/acc crowd, and nearly everyone else out there who discusses AI with insufficient self-consciousness.

Check it out! 🐝

[Contain @con_tain]

New episode 168. with @realityspammer

on the bizarre world of X-Risk, @harmlessai, Corporate Surrealism, E/ACC b*llsh*t, + more, up in bio