The more I learn about AI, the more alarmed I become about its use and implications for society. I think that the current LLM technology we regard -- somewhat erroneously, as I understand it -- as "artificial intelligence" is completely unethical in its design, and is bound to fail, but probably not until it does a considerable amount of harm.
For my own sake, I avoid intentionally using any form of AI at all costs, I won't publish anything with any AI-generated content in it -- and will publicly call out those who do so without disclosing that necessary detail -- and I sure as hell will not spend any money on it. Others may engage with AI applications as they see fit, but understand that choice will be reflected in my judgment of those others' credibility.
I think the AI bubble will burst this year or next, done in by the energy crunch and the weird legal cases that are going to appear with increasing frequency, taking AI developers to task for AI's environmental costs and the manner in which they use data. Those risks are going to make it difficult to obtain funding or insurance, and some developers are going to find themselves underwater. The big ones, Alphabet, Meta, Amazon, etc. will be fine, but most of the speculative 3rd party data center operators are going to be left out to dry. It will be interesting to see what happens then.
(Image: One of the data centers around Dublin. I'm not sure which one, this may be Microsoft's.)
No comments:
Post a Comment