Mastodawn

This is an excellent paper (shared with me by Emily Bender - I am not atting her in order to avoid annoying her further).

It exposes the motivations behind AI development and criticizes the focus on AGI. I thought AGI was just bullshit they used in presentations to sell what they are doing to the public. I mean AGI is obviously not what they are actually developing. "AI" is just ways to analyze data, create models, and create tools on top of this. But apparently they believe their own bullshit.

https://firstmonday.org/ojs/index.php/fm/article/view/13636/11606

View of The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence

Show thread

Aaron 2d ago

The ground is laid in the paper for the following assertion:

"The AGI race is not an inevitable, unstoppable march towards technological progress, grounded in careful scientific and engineering principles. It is a movement created by adherents of the TESCREAL bundle seeking to “safeguard humanity” by, in Altman’s words, building a “magic intelligence in the sky”, just like their first-wave eugenicist predecessors who thought they could “perfect” the “human-stock” through selective breeding."

I agree.

Show thread

Aaron

I thought Altman was just playing Musk (who's ideas are mostly just fuzzy unserious variations of woo woo sci-fi bullshit) in order to get money out of him for the founding of OpenAI. I thought that that was what all the talk about saving us from potential apocalypse driven by hostile AIs was all about. Just snake oil to get the ball rolling on investment.

But there's significant funding and organization behind research, indoctrination, education, and funneling talent toward developing AGI.

Whether they are afraid that AGI could happen somewhere and be weaponized or they are genuine believers that it will be a good thing for humanity and just want to ensure that they profit... either way they believe in it.

Did these idiots have parasocial relationships with the fucking ELIZA program as kids? What is wrong with them?

Show thread

Aaron 2d ago

The problem with this (read the paper to get a deeper understanding of this) is that AGI is not a well defined goal.

Pursuit of AGI results in nightmarish levels of scope creep.

Due to the nature of the tools being developed this is a bigger problem than just painful software development.

All kinds of toxic ideological bullshit joins that scope creep and gets embedded deep in the models.

And I would like to add that on top of that the uses to which people imagine these tools can be applied also include everything under the sun which just drives the potential for even more madness finding its way into computer models that are supposed to mimic human thought processes.

Show thread

Aaron 2d ago

I read this paper because I disagree that "AI" is based in race science. I think it influences the current development of it and for me that is an important distinction.

This paper does not support the assertion that AI is based in race science. BUT if you feel attacked by the assertion that AI is based in race science, you really should read this paper and take the time to digest what it says.