The Real World is Bailing on Artificial Intelligence
EXPERT OPINION BY HOWARD
TULLMAN, GENERAL MANAGING PARTNER, G2T3V AND CHICAGO HIGH TECH
INVESTORS@HOWARDTULLMAN1
Major non-tech public
companies in the U.S. have pretty much squeezed all of the good news and PR
juice possible out of the A.I. lemon. The end of the hoopla around large
language models -- which we once called “vaporware” -- is drawing near.
Not surprisingly the
Potemkin projects these businesses hastily erected as shiny facades to hide the
sad reality that they actually had no clue about what was coming down the pipe
and are already showing their age and their cracks. The recent Sports
Illustrated debacle where fake articles created by A.I. were
published under false names is a good example of pretenders
getting ahead of the curve. None of these folks really had any business trying
to get into this messy and complicated business. They also had no real desire
to make the long-term and expensive investments required to seriously find out
what was going on. Only the Magnificent Seven tech giants can actually afford
to unreservedly pursue pipe dreams on this scale. The rest of these guys were
mostly kidding themselves and trying to play catchup in terms of the market
buzz.
The relatively prudent
and financially responsible companies among them - notwithstanding their latest
A.I. excesses - have for years been shortchanging R&D to bolster their
bottom-line profits. This is what you’d expect, especially when a new area
of invention and innovation is outside of their alleged areas of expertise. The
very best strategy they can muster is rarely substantive. It’s almost always
about noise, free-riding, and jumping on the bandwagon whether they know where
it's headed or not and doing it on the cheap to boot.
The current storm, which
is rapidly abating, is just the latest case of believing that, if the story is
big enough and the shouting is loud enough, the facts don’t matter. At the
moment, they’re done fishing and starting to focus on cutting bait and bagging
the whole show.
Expect that in the first
quarter of this year we’ll see an unending parade of delays, deflections and
dodges all designed to try to let the air out of this particular balloon before
it completely pops. I always thought that some of the craziness around the
diversity, equity and inclusion nonsense and the pronoun police would hit the skids
before the big guys started to bail out of their A.I. misadventures. But it
looks like we’ll be stuck with the politically correct know-it-alls for a while
longer although certain DEI-cheerleading college presidents won’t be leading
the charge. I guess the new rule these days - courtesy of the Orange Monster -
is that you can lie as much as you want, as long as you can get away with it,
and as long as you don’t come to believe your own lies.
This is not to say that
every company won’t eventually benefit from the incorporation of these ideas
and new technologies into the software, operating systems, and other
productivity tools which they purchase and license from the tech giants. It’s
just that all their creatively cosmetic attempts to do any of this magic
themselves are already clearly doomed. Better than 90% of the businesses in the
U.S. may never have any material uses or direct application
for artificial intelligence in their shops.
But that didn’t keep all
the biggest non-tech players from dreaming up, staffing up, launching at great
expense, and then hyping the daylights out of their own A.I. initiatives long
before there was any demonstration of practical applications, productivity
gains, or even validation of the accuracy of the underlying statements, which
have regularly been shown to include what the industry politely calls
“hallucinations.” In formulations and fantasies that Lewis Carroll would be
proud of, the promoters like to suggest that what the systems say is very clear
- what the verbiage means is up to you - and you’re invited to take and use it
at your peril.
The biggest problem with
these fantasies is that they never deliver the goods. Even the most industrious
group of highly paid monkeys sitting in front of a bank of computers won’t
write critical code that is worth anything if they have no idea of what they’re
trying to do. Of course, the people charged with explaining to their finance
teams and eventually to their investors (not to mention the
about-to-be-unemployed team members) where these orphaned A.I. experiments and
abandoned initiatives now stand have their hands full. Because there’s no
amount of creative math sufficient to conceal the true costs of fake work.
You had hundreds of
people charging full speed ahead in these companies under the delusion that
effort alone (rather than vision or intelligence) was sufficient to create
value. Now the marketing and PR efforts will turn to crafting convenient
justifications and explanations - just as baseless and fantastical as the
original story lines - to escape the thorny and inescapable truth that the
money is gone, and they have nothing to show for it. Even ChatGPT has no
answers for this mess. You can either have results, or excuses, but not both.
As all these companies
begin to slow down and back away from their stalled and stale “projects”, I
assume that they’ll be seizing upon a couple of obvious and easy explanations -
as long as it doesn’t highlight their own ignorance. You can be sure that no
one will ask why they didn’t think about these exposures before they rushed
ahead.
We’ll hear about
elaborate concerns over privacy and data. Angst about security and safety, especially
in terms of other countries and bad actors. New issues regarding copyright laws
and, of course, all the scams and frauds based on voice and video simulations.
Measurement, validation and tracking problems will be highlighted as they all
begin to worry anew about accountability. Keep in mind that these are the same
folks spending billions in ad dollars for programmatic ad solutions where they
have absolutely no idea of how the ads are being placed or their impact and
effectiveness.
The bottom line is
pretty simple and two lessons to keep in mind as the new year begins. You can’t
go wrong by sticking to your knitting and focusing on
doing a better job on the things you know how to do. And don’t try to do
complicated things cheaply that you shouldn’t do at all.
JAN 9, 2024
The opinions expressed here by Inc.com
columnists are their own, not those of Inc.com.