More than meets the AI: Past is prescient

Today’s issue is brought to you by: it’s Monday and I made it out of bed, isn’t that good enough?

In this issue:

  • STEMPH (252 words)
  • Pro bono publico (288)
  • I need money (179 words) 

STEMPH

Here’s an argument someone should have made a few centuries ago: history and philosophy should be included in the STEM fields.

But, when I say “history” and “philosophy,” I don’t mean the pop history and junk philosophy that passes for academia in many places. I’m talking about a robust and complex understanding of precedence and humanity that’s as dependable and useful as the study of biology, chemistry, computer science, or physics.

For example, if you’d like to gain a greater understanding of the current “ChatGPT and Microsoft” zeitgeist, you might want to study up on the relationship between Igor Sikorsky and Henry Ford circa the late 1930s and early 1940s. 

I also think even a cursory philosophical analysis of how Meta chose to develop, market, and launch its Galactica chatbot model can explain its abject failure and subsequent public recall.

Being a code genius will only get you so far in the modern world. The ability to memorize facts or snippets of information is far less impressive today than it was back when we needed a library pass and access to microfiche to verify facts.

In other words: analysts are more important now than ever. 

The bottom line here is that STEM companies need futurists. But those futurists should be far more comfortable reciting history or applying philosophy than they are at “predicting the future.”

Nobody can predict the future. Full stop. 

What we can do is study humanity, learn from the past, and apply those lessons to the future. 

Pro bono publico

In service to the public interest, I recently analyzed Meta’s “Galactica” model. It’s an LLM that, according to its accompanying paper, was trained on science papers, Wikipedia, and other data. Supposedly, it was meant to be an assistant or research aid for scientists to help them generate papers at the push of a button.

What it actually does, in its most recent public-facing incarnation, is act as an invitation for political snake oil peddlers to distribute misinformation.

That’s not what Meta intended, but good intentions don’t affect human outcomes as much as those who end up defending them would like everyone to believe.

What I do believe, however, is that had Meta actually paid attention to similar launches in the past, it might not have found itself in the unenviable position of being responsible for the biggest public chatbot flop since Microsoft launched Tay

You’d think Yann LeCun and the FAIR gang would have studied that launch and figured out what went wrong (hint: nobody needs a public-facing text generator whose outputs can be redefined in real-time by the worst people on the internet).

And, if they had, they might have understood that launching a novel, untested, brittle chatbot to the general public and claiming it’s a “TOOL FOR SCIENTISTS” is a lot like selling candy-flavored cigarettes with cartoon mascots printed on the packs and claiming they’re ONLY FOR ADULTS.

My point is: the intent is irrelevant when the outcome is so incredibly predictable. There will always be some people who think that the claims made about a product are all that really matter. 

Those people are typically well-versed in neither history nor philosophy. You can usually find them reacting poorly to bad news. 

I need money

Your love give me such a thrill

But your love don’t pay my bills

I need money (That’s what I want)

That’s what I want (That’s what I want)

That’s what I want (That’s what I want)

That’s what I want (That’s what I want)

Just some quick, shameless self-promotion in the form of an FYI: I’m looking for work. I was laid off right before the holidays and I’m still trying to find my next challenge (and a steady paycheck to feed my family!).

I’m available for work as a journalist, writer, editor, speaker, consultant, and futurist. I’m currently working freelance (and hosted by Loveisafetish.com), but I’m open to contract and staff positions. I’d love to speak at your event, write an op-ed, newsletter, or column for your publication, or even just hop on a call to discuss the future.

I can be reached at: tristangreene1977@gmail.com or on Twitter.

If you’d be so kind as to share this information with your social and employment networks, I’ll forever be in your debt.

via GIPHY

 Read more More Than Meets the AI here!

Authors

We would love to hear your thoughts!

5 reasons why you should start a mini water propagation garden today