In late July, I warned about a large tech company riding on the “AI Crazy Train.”
Recently, I introduced an Enron documentary to my daughter. And once again I was shocked at how much the company begged Wall Street as it pursued wild and expensive infrastructure projects.
These two threads came together in the form of a fascinating interview I just discovered. The new Institute of Economic Thinking recently won an interview with renowned shortseller Jim Chanos.
Dr. Lynn Parramore, senior research analyst at INET, interviewed Chanos on many topics, and his comments on AI really stood out to me.
Chanos is a well-known short seller. He ran a hedge fund company called Kynikos Associates, which focused most on betting on stocks. This is very difficult and he did it pretty well.
His most famous trade was to shorten Enron's inventory before the energy trading giant collapsed in 2002 with a massive bankruptcy.
Yes, Enron was an energy company, but he tried to exchange internet bandwidth. Internet bandwidth was intertwined with communications and broadband giants such as MCI Worldcom, which was bustled even in the collapse of the dot-com in the early 2000s.
Chanos closed his hedge fund company and transformed his business into a family office several years ago. However, he is worth listening, so check out the excerpts from a recent interview with INET below.
Related Stories
Business Insider tells the innovative stories you want to know
Business Insider tells the innovative stories you want to know
(Through INET, I asked Chanos if they were shorting up any of the companies involved in this AI spending frenzy. Chanos said he wasn't talking about the company in the name of the interview.
Possibility of AI to destroy old business models:
“I don't know where AI will go and take us. By this month, it's getting better each year. We have to look. I'm as interested as anyone – replace a particular job?
The Internet was an incredible development. Basically, we were able to incorporate analog digitalised items, transfer them almost for free, and store them for free. It has created all sorts of new businesses. It also destroyed many old businesses. Businesses that had digitally made analog products were getting into trouble, like Eastman Kodak.
I think we're going to see the same thing with AI, so we need to see how it works. ”
About the oddity that new technologies can't increase productivity:
“But one more interesting takeout from the dot-com era and the networking boom is seeing our real GDP ten years before Netscape from 1987 to 1997 when we all started communicating with each other.
Therefore, there was no increase in economic growth. Productivity increased somewhat, but due to low population growth, total economic growth did not increase significantly. Overall, the boom we expected to see in the world's most advanced economy was not noticeable because of the internet. And that hasn't been since. It would be interesting to see if AI would change that. ”
How accounting drives huge AI spending
“These things, and certainly we did with fracking, made them ominous in the Telecom and the.com era – drive capital spending and revenue for very interesting reasons. It's not just agitated with the animal spirit, there's also an accounting angle that everyone forgets. There's no underlying return.
What we're currently seeing with AI chips is similar to what happened during the communications boom. Large tech companies, or so-called Mag 7 – are buying tips as quickly as possible. A company that sells tip books that spend as income and profits, like Lucent and Nortel did at the time. However, companies that have previously spent such as MCI and AT&T are those who are paying for the costs. They capitalize their costs and spread amortization over several years. Therefore, chips that can become obsolete in just two years are depreciated by large hyperscalers of five, six, and even seven years or more. It will significantly boost corporate revenues during the high-tech build-out boom that we see right now. That's the best. ”
What happens when the AI spending boom stops?
“Second, spending, and therefore revenue, can collapse. Everyone remembers the global financial crisis, but forgets that between 2000 and 2002, corporate revenue fell by 40%, just like the global financial crisis.
I'm worried that there will be a lot of spending right now in the physical boom in AI (build-outs of data centers, chips, etc.). That could be a big problem. When we spend $50 billion a year, that's one thing. When it changes, it's something else. As a group, we spend $500 billion a year. Without actual returns, can we spend $1 trillion each year on this without looking at the results? It would be interesting to watch. Expenses are currently increasing faster than operating profit or revenue. ”
An unpleasant question will come soon
“So, within a year or two, we have now reached the point where some of these big companies have to make very uncomfortable decisions about when and how to monetize AI, and ultimately make very uncomfortable decisions about how this massive spending will turn out.
Sign up for BI's Tech Memo Newsletter here. Please contact me by email abarr@businessinsider.com.
