Artificial intelligence. Last year I shared a prediction by Gartner that 85% of AI business solutions would fail by 2022.
With a wave of recent tech layoffs, it seems unlikely the tech giants will have a massive tech breakthrough. They’re preparing for tough times, which signals to me they’re probably not expecting anything revolutionary. According to some sources, as many as 73,000 tech workers have been laid off this year.
As reported by Fortune and Kevin Kelleher, Deloitte has released a report. A survey of 2,620 executives in 13 countries found 94% of respondents considered AI critical to success in the coming five years. Yet only about 50% of solutions were considered high achieving, leaving about 50% of solutions low achieving.
The article does state that global spending on AI is set to nearly double from $33 billion in 2021 to $64 billion in 2025, but this does not seem to comport with the dropping interest in AI investment reported. Only 76% of respondents plan to boost AI investments, down from 85% a year ago. As I see it, there are two possible futures. Either executives will take a step back and realize that somewhere between half and 80% of these projects are not working out, or they’ll continue down the path of burning money on what’s essentially a hand of blackjack with a nice side bet. For those that don’t gamble, that’s “okay odds of winning something with a very small chance of making a lot of money.”
Of course, this is all about AI generally. Automatic speech recognition (ASR) or natural language processing (NLP) is the subset of AI that deals with or theoretically threatens our profession, and if regular AI solutions are not working out, something as complex as ASR doesn’t seem like it’ll be perfected any time soon.
We are in a bad time for factual information. Some sources claim 15 to 30% of market research data is fraudulent. I can’t help but wonder if the ballooning estimates on market size for AI and ASR are wishful thinking put out there to get investors and decision makers to part with their money and avoid another AI winter. After all, it looks attractive to say speech recognition will grow from $11 billion in 2022 to $49 billion in 2029, and it looks attractive to say the AI market will grow to $1.3 trillion by 2029. “Buy in now!” But if half the solutions are low achieving, aren’t they throwing half of that trillion in the trash? And isn’t it a little odd to forecast that an industry will miraculously grow from $64 billion in 2025 to $1.3 trillion in 2029? Some of the fastest-growing industries are around 12% growth. Meanwhile they’re talking about something like 500% growth every year for 4 years. I know jobs don’t necessarily relate to revenue/profit, and I accept these figures are from different sources and therefore may be calculated differently, but how is such explosive growth expected to take place without adding jobs? Were the thousands of workers just laid off not contributing to the growth? Would the growth not slow with such future uncertainty? The AI industry is set for 500% growth but the world GDP is something like 10% growth in a great year?
This also calls into question the market research and data in our field. Will our “sten tech” companies’ AI ventures be successful? Could numbers be played with to get court reporting consumers to buy into new tech ideas? Could the numbers be played with to make court reporting firms more attractive to private equity buyers? I’m already certain they’re being played with to sell digital court reporting.
I will continue to document the stats. Let this serve as a reminder that we need to be careful what stats we accept in our minds as true and always be open to new information. As always, I invite readers to share comments, contrary ideas, or even correct me. What do you think?
Addendum:
Tech giant Amazon basically admits mileage varies depending on who’s speaking.