Is the AI Industry About to Collapse?
Thad McIlroy
New book: "The AI Revolution in Book Publishing: A Concise Guide to Navigating Artificial Intelligence for Writers and Publishers"
This past week has been rough on AI boosters like myself. I see the many threats potentially posed by generative AI, and acknowledge those in my book . But it's not the threats that caused the biggest stir this past week. It's that the whole AI "house of cards" may be about to collapse.
Tomorrow, Monday, will be an interesting day at the stock markets. Last week was tough on the share prices of the big AI-connected players, Google, Microsoft, Nvidia, et al. I see, late Sunday night, in the Wall Street Journal "Japan’s Nikkei Posts Biggest Single-Day Fall Since 1987 After Weak U.S. Data."
The stock market is such an imperfect mirror of business reality, but it's always seen as a proxy, and if the Magnificent Seven share prices continue to drop, the overall mood surrounding generative AI is going to take a big hit.
I look into this further in a new blog post .
Visionary AI-First Executive Driving Business Innovation | Expert in SaaS, Media, and Digital Strategy | Leader in Growth and Operational Excellence | Artificial Intelligence Certifications
3 个月Very insightful, as always, Thad McIlroy. Yes, the stock market has been extremely volatile lately, but I think we all suspected a correction was on the horizon. And this won't be the last one, of course. As someone smarter than me once said, "the stock market is not the economy and the economy is not the stock market." According to Gartner's Hype Cycle, we're in the trough of disillusionment stage for AI. It happens with every new technology and AI's widespread buzz over the past 2+ years makes this a deep trough. Expectations are largely to blame for much of the disillusion. As another wise person recently said, "we were promised flying cars and they gave us automated email templates." Another factor leading to disillusion is the misguided assumption that AI requires a data scientist and a team of software engineers. The AI development we've been doing here at Revenue Path Group featured neither of those thanks to the power and flexibility of a no-code environment. Our investment up to now is in the low tens of thousands of dollars, 95% of which is simply my time, and we have a solution we recently started presenting to clients. Every new tech has a shakeout period but AI, including generative AI, is here to stay, IMHO.
I agree with Marcus on a lot, but he's got thin skin and personalizes things when he should not. He is not, however, anti-AI. I'm probably not telling you anything you don't know, but, in case others don't, he believes is that the pure neural-network architecture of current LLMs are an impressive dead end, incapable of knowing or reasoning about anything. He demonstrates again and again that they can't perform the most basic reasoning tasks, but merely follow statistical grooves that seem like it, under the right conditions. (Have you read Rebooting AI?) His hope for much better AI lies in adding a strong symbolic component—an AI that has structured knowledge and structured reasoning, not conceiving of these as somehow "emergent" from running a lot of content through a neural network until it can reproduce it accurately. FWIW, I suspect he's right that current LLM architectures aren't going to get us "real" reasoning. At the same time, they're useful for a lot of things, and have at least some room to improve. Does that justify all the current spending? Unclear. But, as I said, he personalizes things too much. And like many a voice in the wilderness, he's prone to using extreme language just to be heard at all.
Oh come on… this is not about AI at all. This is about the financial markets! AI is obviously overhyped, because that’s what the markets always do. But that won’t affect the long-term real value of AI. This is the same cycle that has happened repeatedly in tech. Heck, it happened to AI in the early 90s! (How many technologies get to be overhyped into a bubble twice, 30 years apart?!)