Dr. Damien P. Williams: Magus, Werewolf, Cyborg, Bi: So. Google just released an internal paper about the epistemic, ethical, and sociopolitical threats of generative "AI," and golly gee whiz if that doesn't sure as shit sound familiar. 🤔🧐😒🙄 www.404media.co/google-ai-po...

So. Google just released an internal paper about the epistemic, ethical, and sociopolitical threats of generative "AI," and golly gee whiz if that doesn't sure as shit sound familiar. 🤔🧐😒🙄 www.404media.co/google-ai-po...

Google: AI Potentially Breaking Reality Is a Feature Not a Bugwww.404media.co “While these uses of GenAI are often neither overtly malicious nor explicitly violate these tools’ content policies or terms of services, their potential for harm is significant.”

Literally admits that "hallucinations" and bias are "limitations of GenAI systems themselves." My god 😂😭

Just Like I Fucken Told You They Were youtu.be/9DpM_TXq2ws www.americanscientist.org/article/bias... bsky.app/profile/wolv...

Literally admits that "hallucinations" and bias are "limitations of GenAI systems themselves." My god 😂😭

Argh, they are not hallucinations 😫, those mofos link.springer.com/article/10.1...

ChatGPT is bullshit - Ethics and Information Technologylink.springer.com Recently, there has been considerable interest in large language models: machine learning systems which produce human-like text and dialogue. Applications of these systems have been plagued by persist...

Oh I'm well aware youtu.be/9DpM_TXq2ws

On Bullshit Engines: The Socioethical and Epistemic Status of GPTs and other AI. Dr. Williams.youtu.be Forum for Law Ethics and Society (FiLES). Dr. Williams takes a deep dive into the propensity for GPT style AIs to not merely present falsehoods but to perpetuate bias, stereotypes, and other BS as if it was legitimate. @SonomaStatePhilosophy

They are also limits of natural intelligence systems, aka, people. In my experience, I am no more likely to get a correct answer from people than from AI. EG, ask someone about the existence of a super cloud daddy in the sky.

You sit in a relational, interconnected, mutually sustaining matrix with other intentioned beings, and if "AI" were built in a completely different way, it might be able to be that on purpose, but it is *not* built that way, and every intention it currently has is an intention built from capitalism…

…militarism, hegemony, and white supremacy, & it has neither the inclination nor the capacity to meaningfully introspect and bracket those biases. In short, even if this form of "AI" *were* a person, it's for sure not a person i want influencing every single aspect of my daily life. Have a good one.

AI, like people, is a variegated category. For AI, we don't even know what it is, really, since it's such early days. But I do agree that I do not want AI influencing me all the time but that applies to corps generally. I paid Kagi for search for exactly that reason, well before Google added AI.

The thing is, I know lots of people who seem to lack the "inclination[or] the capacity to meaningfully introspect and bracket those biases". I have a client, for example, who fits that, along with some of the other negative characteristics. Still, he is a generally pleasant person who pays on time.

I like PI.ai, for example, and think it has great potential for helping people who are isolated or need a dispassionate context to vent their ideas, etc. Were you to day, "I think ChatGPT and Gemini are manipulative corporate agents", I would likely agree, though perhaps not about the specifics.

Pi, your personal AIpi.ai Hi, I'm Pi. I'm your personal AI, designed to be supportive, smart, and there for you anytime. Ask me for advice, for answers, or let's talk about whatever's on your mind.

I think denigrating AI generally is incorrect generalization. Any of them can be bad, or good. The tech, as far as I can tell, and I am an enthusiastic dilettante, is value neutral. Training data, fine tuning, human feedback, and other stuff matter a lot.

I mean… I don’t even know what to say here. We’re fucked

Dead Internet Theory to be real at TBD 😂

And yet the company continues to do R&D into generative AI like good greedy capitalists

Worried but not worried enough to avoid profit

Given the energy cost vs. lack of consumer acceptance, continued investment is a good way to avoid profit.

Truth

What are the most important points?

"Namely, that generative AI makes it very easy for anyone to flood the internet with generated text, audio, images, and videos." My understanding is basic. I think it means access to AI tech means access to the proliferation of what we typically identify as computer viruses. Clogging systems w/data

Except instead of simple logic loops et al. We're just hard producing an unfathomable amount of data into servers. From an ecological stand point I know that this is significant stress on our power consumption.

The paper is accessible through this link by OP. It's about 29 pages it looks like ive got lunch reading

tbf to have much chance of avoiding that you’d need a training corpus scrubbed free of most corporate & executive behavior…which would be a monumental task….

Yep.

Post