Most of this Ted Gioia post is behind the paywall (and if you can afford it, I’m sure you’d get your money’s worth for a subscription):
I anticipate extreme turbulence on every front for the remaining five months in 2024. You will see it in politics, business, economics, culture, world affairs, the stock market, and maybe even your own neighborhood.
That’s one of the themes of my latest arts and culture update below.
What happened to the AI business model last week?
After almost two years of hype, the media changed its opinion on AI last week.
All of a sudden, news articles about AI went sour like reheated 7-Eleven coffee. The next generation AI chips are delayed, and 70% of companies are behind in their AI plans. There are good reasons for this — most workers now say AI makes them less productive.
People are also noticing that AI businesses want to use the entire electricity grid to run their money-losing bots. Meanwhile AI companies are burning through cash at historic levels. Even under the best case scenario, this all feels unsustainable.
But the worst disclosure, in my opinion, came on July 24 — just eleven days ago.
A study published in Nature showed that when AI inputs are used to train AI, the results collapse into gibberish.
This is a huge issue. AI garbage is now everywhere in the culture, and most of it undisclosed. So there’s no way that AI companies can remove it from future training inputs.
They are caught in the doom loop I described last week.
That same day, the Chief Investment Officer at Morgan Stanley warned investors that AI “hasn’t really driven revenues and earnings anywhere”. One day later, Goldman Sachs quietly released a report admitting that the AI business model was in serious trouble.
Even consulting firms, who make a bundle hyping this tech, are backtracking. Bain recently shared the following chart (hidden away at the end of a report) which explains why AI projects have failed.
These findings are revealing. They show that management is absolutely committed to AI, but the tools just don’t deliver.
And, finally, last week the media noticed all this.
They published dozens of panic-stricken articles. Investors got spooked too — shifting from greed to fear in a New York minute. Over the course of just two days, Nvidia’s stock lost around $400 billion in market capitalization.
In this environment, true believers quickly turn into skeptics. The whole AI business model gets scrutinized — and if it doesn’t hold up, investment cash flow dries up very quickly.
This is exactly what I predicted 6 months ago. Or even a year ago.
I expect that the next few weeks — or maybe even the next few days — will be extremely turbulent in the AI world.
Buckle up!
The dominant AI music company just admitted that it trained its bot on “essentially all music files on the Internet”.
Suno is a huge player in AI music — it tells investors it will generate $120 billion per year. Microsoft is already using its technology.
But there’s a tiny catch.
The company now admits in a court filing:
Suno’s training data includes essentially all music files of reasonable quality that are accessible on the open internet, abiding by paywalls, password protections, and the like, combined with similarly available text descriptions
Hey, this is totally illegal — it’s like Napster all over again.
Suno will need to prove that all these copyrighted songs are “fair use” in AI training. I doubt that any court will take that claim seriously.
If the music industry is smart, they will use this violation to shut down AI regurgitation of copyrighted songs.
If the music industry is stupid — run according to my “idiot nephew theory” — they will drop charges in exchange for some quick cash.