iiinova — where your digital self keeps evolving

Rational Animations @rationalanimations.bsky.social · 21/09/2026

Can we reduce global risks without enabling totalitarianism?

Rational Animations @rationalanimations.bsky.social · 20/09/2026

In this new video, we explore warning shots: events that could wake humanity up to the dangers of AI. We ask whether we'll get such a warning, and what past examples involving nuclear power and pandemics can teach us about how people might respond: youtu.be/Y2vDwR9rW-0

youtu.be

What will it take to wake humanity up?

In this video, we explore warning shots: events that could wake humanity up to the dangers of AI. We ask whether we'll get such a warning, and what past examples involving nuclear power and pandemics…

010

Rational Animations @rationalanimations.bsky.social · 19/09/2026

Can we prevent technology from being abused without outlawing it?

000

Rational Animations @rationalanimations.bsky.social · 17/09/2026

Can regulation help against totalitarianism?

000

Rational Animations @rationalanimations.bsky.social · 15/09/2026

Could Mind-Reading Technology Empower Tyrants?

000

Rational Animations @rationalanimations.bsky.social · 13/09/2026

What Technologies Could Empower Dictatorships?

021

Rational Animations @rationalanimations.bsky.social · 11/09/2026

Could totalitarianism be on par with human extinction?

011

Rational Animations @rationalanimations.bsky.social · 09/09/2026

Could an Evil Regime Rule the World Forever?

010

Rational Animations @rationalanimations.bsky.social · 22/08/2026

Could a totalitarian regime conquer the entire world and lock in its power forever? In this new video, we explore which technologies might make this possibility more likely, and how we could shape technological development in order to prevent it: youtu.be/2Wv3p9WCs6M

youtu.be

Could a Global Dictatorship Last Forever?

Could a totalitarian regime conquer the entire world and lock in its power forever? This is a particular type of unrecoverable dystopia: a future in which humanity survives, but forever loses its…

172

Rational Animations @rationalanimations.bsky.social · 09/08/2026

Rogue OpenAI models hacked Hugging Face to cheat on a test. Here's our recreation of the event!

160

Rational Animations @rationalanimations.bsky.social · 29/07/2026

AI is more and more capable of original thinking:

000

Rational Animations @rationalanimations.bsky.social · 26/07/2026

Since you're all fans of Ada Lovelace, we thought we'd give you a sneak peek behind the scenes. Here's some pre-production artwork from when our art director, @peonyvibes.bsky.social, was cooking up her design! Ada Lovelace appears in our latest video. You can watch it here: youtu.be/Z7dH_Cxx75g

010

Rational Animations @rationalanimations.bsky.social · 23/07/2026

Can computers come up with original ideas?

000

Rational Animations @rationalanimations.bsky.social · 18/07/2026

Hello hello! You should know that Rational Animations is a US 501(c)(3) nonprofit. This means that if you're a US taxpayer, direct donations are generally tax-deductible! We recommend supporting us directly through our website: www.rationalanimations.com/donate

010

Rational Animations @rationalanimations.bsky.social · 17/07/2026

AI is often dismissed as just a word predictor that's incapable of original thought. Is this view correct? In our new video we try to answer by looking at the history of AI, modern interpretability studies, and the way current AI is actually trained to solve complex problems.

youtu.be

- YouTube

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

021

Rational Animations @rationalanimations.bsky.social · 07/07/2026

How one typo sent GPT-2 spiraling into a smut-maximizing machine:

010

Rational Animations @rationalanimations.bsky.social · 06/07/2026

The alignment training that made GPT-2 maximally lewd:

020

Rational Animations @rationalanimations.bsky.social · 03/07/2026

The origin story of the lewdest AI ever created:

000

Rational Animations @rationalanimations.bsky.social · 16/06/2026

Are AIs ironically hampering our safety tests?

010

Rational Animations @rationalanimations.bsky.social · 15/06/2026

AIs know when they're being tested for safety

020

Rational Animations @rationalanimations.bsky.social · 14/06/2026

Here's what happened when researchers tried to make an AI model scheme less:

000

Rational Animations @rationalanimations.bsky.social · 13/06/2026

Why an AI model hid its true capabilities from researchers:

010

Rational Animations @rationalanimations.bsky.social · 06/06/2026

AI Safety researchers found that, in certain situations, AI models can take covert actions. Additionally, they're sometimes aware they're being tested, which causes them to behave better. Our new video discusses these results and more. You can watch it here: youtu.be/hzlR0R91lZA

youtu.be

AIs don't scheme... if they know we're watching

YouTube video by Rational Animations

031

Rational Animations @rationalanimations.bsky.social · 04/06/2026

How a deceptively aligned AI looks perfectly fine until it takes over:

011

Rational Animations @rationalanimations.bsky.social · 03/06/2026

How human drives decorrelated from reproductive success:

010

Rational Animations @rationalanimations.bsky.social · 02/06/2026

AI capabilities can generalize while goals remain misaligned:

010

Rational Animations @rationalanimations.bsky.social · 01/06/2026

How a tiny shift in the environment broke an AI agent:

000

Rational Animations @rationalanimations.bsky.social · 30/05/2026

Humans are misaligned from the perspective of evolution:

020

Rational Animations @rationalanimations.bsky.social · 18/05/2026

Here's how we broke out of the simulation:

010

Rational Animations @rationalanimations.bsky.social · 18/05/2026

They were not aliens. They were something much weirder.

042

Rational Animations @rationalanimations.bsky.social · 17/05/2026

It looks like our simulators aren't very bright...

000

Rational Animations @rationalanimations.bsky.social · 15/05/2026

It looks like aliens are sending us a picture.

111

Rational Animations @rationalanimations.bsky.social · 14/05/2026

Someone is talking to us by... shifting the luminosity of stars?

000

Rational Animations @rationalanimations.bsky.social · 12/05/2026

Here's why security measures won't work on superintelligence:

110

Rational Animations @rationalanimations.bsky.social · 06/05/2026

We might only get one real attempt at superintelligence

000

Rational Animations @rationalanimations.bsky.social · 02/05/2026

Developing a superintelligent AI that does what we want without killing everyone may be extremely difficult. In this video, we explain why, using arguments from "If Anyone Builds It, Everyone Dies" by Eliezer Yudkowsky and Nate Soares: youtu.be/7wy3xyoXYt8