Is A.I. Just Another Big Investor Head Fake?

I was one of the progenitors of viable generative A.I. back in the early 2010s. At Automated Insights, we produced machine-written, data-rich articles at the rate of thousands of articles per second for customers like Yahoo Fantasy Football and the Associated Press.

We exited to a private equity firm in 2015.

Since those long-ago heady days, I’ve been mildly impressed but mostly meh on this new GPT-driven wave of A.I. I get it, and I think there are some great applications out there. Lots of potential. But does it really deserve the fervor and, more important, the capital being poured into this somewhat new and novel tech? And why now?

It’s a question that slapped me in the face again recently. But it took me a while to connect the dots. And again, I’m just asking questions here, not pointing fingers.

Yet.

Oops, We Did a Mad Lib

Generative A.I. recently had a major “oops” moment. And outside of a few chuckles and I-told-you-sos from machine-learning PhDs, the moment just kind of came and went.

Here’s how I learned about it.

Not long ago, I did some consulting for a major financial firm as it began to explore using generative A.I. to make recommendations to customers. This was not your basic “My ChatGPT Buddy” or a new version of telephone support’s “press 1 to choose your own personal hell.” This was real people and their real money. 

We spent a lot of time on personalization and algorithms and coding the machine to think like a writer–remembering what had already been said, building on that, and so on.

Anyway, I was on a call with one of the folks I had worked with at that firm, and he brought to my attention the recent failure of Gannett to hide the source code of its version of generative A.I. 

From the article, “The Worthington Christian [[WINNING_TEAM_MASCOT]] defeated the Westerville North [[LOSING_TEAM_MASCOT]] 2-1 in an Ohio boys soccer game on Saturday.”

This is not A.I., I thought to myself. This is templates and Mad Libs. This is what I had to make sure we didn’t do–the kind of stuff our competitors were passing off 12 years ago. 

Then I remembered, our competitors got funded too.

When Did You Decide That NFTs Were Bogus?

Indulge me in another fun story. So, about six years ago, a beverage company called the Long Island Iced Tea Corporation formally changed its name to the Long Blockchain Corporation while stating it was exploring changing its business model from beverages to bitcoin. 

Remember that? I remember that. That’s not the kicker. 

I looked the rest up on Wikipedia. Roughly a year after that obvious tomfoolery, in 2018, the Nasdaq threatened to delist Long Blockchain. Then, in 2021, while we were all mired in Covid, you may or may not have learned that Long Blockchain was indeed delisted.

That’s not the kicker.

Shortly after that, the SEC filed charges. 

Now we’re on to something, and here’s the kicker.

During what was being sold as one of the largest potential technological evolutions in history — that of the complete decentralization of information, currency, and even things — “the effect [of the name change] on [Long Blockchain]’s stock price was almost instantaneous. Shares spiked as much as 500 percent in premarket trading, eventually settling at a 275 percent gain.”

Those SEC charges were filed for “insider trading against three major Long Blockchain investors who allegedly bought substantial numbers of shares that they sold after the stock gained as much as 380 percent. They allegedly had advance notice of the name change, which preceded and caused the stock rise.”

Now, that’s just history from Wikipedia. 

The shocker is how much legitimate investment capital has continued to be poured into blockchain, crypto, and NFTs, through and after 2021. 

The correlation I’m making is how many companies still come to me for help with their amazing new blockchain-powered startup idea. And how many companies have started to come to me for help with their amazing new A.I.-powered startup idea.

And how similar those ideas are starting to sound.

[[CLEVER_ARTICLE_SUBHEADING]]

What if the Gannett moment is A.I.’s version of the Long Blockchain moment?

Yes, I’m hiding behind my question. I’m not saying this is happening. I’m just asking, what would it look like if it did? 

Would it look like the NFT Beeple sold for $69 million in March 2021–four years after Long Island Iced Tea became Long Blockchain?

Would it look like the A.I. front-runner, OpenAI, being valued at $90 billion?

Would it look like four-week-old A.I. company Mistral raising $113 million?

Would it look like Gannett publishing an article with the words “[WINNING TEAM MASCOT]” in the text?

Because all that happened.

I’m Seeing a Pattern, That’s All

I mean, maybe there’s no causation, but the anecdotal data is right there in front of me in my inbox. And if I don’t ask these questions, who will?

Look, I’ve been out of the professional A.I. game for a while now. I made my bones. I’ve got no stake in this science other than the desire to not see another potential game-changing technology turn into a chaotic race to the bottom. 

Well, that and I’ve got some ideas. But I’ll save those for future articles.

I kinda believed in crypto, but not with my wallet, and I never believed in NFTs, but I believe in A.I. I know it’s real because I’ve already lived it. Maybe that’s what makes me cast a more skeptical eye at the “infinite” potential uses for A.I. to better our lives, and makes me hope we all cast a more skeptical eye at that potential before we throw all the money at it.