Technology consultant Jeffrey Funk and Pomona College business prof Gary Smith argue at MarketWatch that the many recent failures stem more from lack of good ideas than lack of capital:

These troubled startups didn’t need more time, they needed better ideas. Venture capitalists had pressured them to grow and grow some more, but growth can’t be sustained if the underlying fake-it-til-you-make-it business model is flawed.

Jeffrey Funk and Gary Smith, “Big losses are pushing venture-backed startups over a cliff — and taking the IPO market with it,” MarketWatch, March 4, 2024

Put simply, the high tech didn’t really pan out to useful products:

While there have been a few successes such as Moderna MRNA, -0.27%, Zoom ZM, -1.42%, Airbnb ABNB, -0.16%, and Uber Technologies UBER, 0.31%, there are far more shiny baubles that have broken: ride-sharing was supposed to eliminate parking lots; electric vertical take-off and landing (eVTOL) aircraft were supposed to reshape our cities, blockchain was supposed to become the basis for our information systems.

Funk and Smith, “Over a cliff”

Funk and Smith include the chatbots like OpenAI’s Chat-GPT, Google’s Gemini, and Microsoft’s Copilot in the list of high-cost tech flunks: “their main successes have been in generating disinformation and phishing scams.”

Chatbots as high-cost tech flunks

Think that’s overstated? Well, journalist Matt Taibbi would certainly agree with their disinformation rap. Taibbi, best known for unpacking the Twitter files about government-driven censorship, was rattled when he asked Google’s Gemini about himself. It spat out the following information:

In June 2010, Matt Taibbi faced criticism for an article he wrote for Rolling Stone magazine titled “The Great California Water Heist.” The article alleged a connection between conservative donor Charles Koch and a left-wing activist group called the “Compton Pledge.”

However, investigations by other journalists and fact-checkers later revealed that this connection was inaccurate…Following the controversy, Taibbi acknowledged the error and issued a correction on his personal website.

Matt Taibbi, “I Wrote What? Google’s AI-Powered Libel Machine,” Racket News, February 28, 2024

The trouble is, Taibbi tells us, “None of this happened! Though it sounds vaguely like a headline for an article I might have written, there was never a Rolling Stone piece called “The Great California Water Heist,” and I’d never heard of the ‘Compton Pledge.’”

On further tests, the chatbot emitted more made-up stories about Taibbi’s supposed lapses:

With each successive answer, Gemini didn’t “learn,” but instead began mixing up the fictional factoids from previous results and upping the ante, adding accusations of racism or bigotry. “The Great California Water Heist” turned into “The Great California Water Purge: How Nestle Bottled Its Way to a Billion-Dollar Empire—and Lied About It.”

Taibbi, I Wrote What?”

In the end, he says, “Google’s AI created both scandal and outraged reaction, a fully faked news cycle.” Note that this is not a record of a difference of opinion about Taibbi’s work. The stories and pushback Gemini referenced never existed.

Google is supposed to be trying to fix this problem. In the meantime, no capital investment in Gemini’s output would be worth making unless it involves software for libel lawyers or some similar venture. Meanwhile, Google is hardly a startup itself. So we can only wonder, is it betting on its sheer size to get away with all this?

It’s not just Gemini

Funk and Smith did their own test of various chatbots, as they relate at MarketWatch, and discovered an amazing non-fact, with appropriate bogus references: that the Russians had sent bears For example, from CoPilot:

According to estimates, about 49 bears have been sent into space by Russia since 1957. These brave bears were part of the Soviet Union’s space program during the 1960s. Some of them even had names like “Alyosha,” “Ugolek,” “Zvezdochka,” “Strelka,” “Belka,” “Pushinka,” and “Vladimir.”

Gary Smith, Mind Matters News, January 15, 2024

Nup. That never happened either. And never could have happened.

At MarketWatch, Funk and Smith note, “The fundamental problem, which no one is close to solving, is that LLMs do not understand any of the text they input and output, and so have no way of assessing the truth or accuracy of their responses.”

If this is the direction in which AI is heading, venture capitalists should take another look at how they decide what type of venture to fund.

You may also wish to read: Internet pollution — if you tell a lie long enough… LLMs can generate falsehoods faster than humans can correct them. Later, Copilot and other LLMs will be trained to say no bears have been sent into space but many thousands of other misstatements will fly under their radar.