https://thedailywtf.com/articles/ai-the-bad-the-worse-and-the-ugly
…the average American, I think, has fewer than three friends. And the
average person has demand for meaningfully more, I think it's like 15
friends or something, right?
— Mark Zuckerberg, presumably to one of his three friends
"Since even the President of the United States is using ChatGPT to cheat on his
homework and make bonkers social media posts these days, we need to have a talk
about AI.
Right now, AI is being shoe-horned into everything, whether or not it makes
sense. To me, it feels like the dotcom boom again. Millipedes.com! Fungus.net!
Business plan? What business plan? Just secure the domain names and crank out
some Super Bowl ads. We'll be RICH!
In fact, it's not just my feeling. The Large Language Model (LLM) OpenAI is
being wildly overvalued and overhyped. It's hard to see how it will generate
more revenue while its offerings remain underwhelming and unreliable in so many
ways. Hallucination, bias, and other fatal flaws make it a non-starter for
businesses like journalism that must have accurate output. Why would anyone
convert to a paid plan? Even if there weren't an income problem—even if every
customer became a paying customer—generative AI's exorbitant operational and
environmental costs are poised to drown whatever revenue and funding they
manage to scrape together.
Lest we think the problem is contained to OpenAPI or LLMs, there's not a single
profitable AI venture out there. And it's largely not helping other companies
to be more profitable, either.
A moment like this requires us to step back, take a deep breath. With sober
curiosity, we gotta explore and understand AI's
true strengths and
weaknesses. More importantly, we have to figure out what we are and aren't
willing to accept from AI, personally and as a society. We need thoughtful
ethics and policies that protect people and the environment. We need strong
laws to prevent the worst abuses. Plenty of us have already been victimized by
the absence of such. For instance, one of my own short stories was used by Meta
without permission to train their AI."
Cheers,
*** Xanni ***
--
mailto:xanni@xanadu.net Andrew Pam
http://xanadu.com.au/ Chief Scientist, Xanadu
https://glasswings.com.au/ Partner, Glass Wings
https://sericyb.com.au/ Manager, Serious Cybernetics
