One of the entrepreneurs hit hard by the dotcom crash of 2000 says there is room for optimism amid fears of the artificial intelligence bubble bursting.
"It's definitely a bubble but it's a different kind of bubble," Tom Herman, co-founder of GovWorks, told The National. GovWorks, founded in 1998, promised to revolutionise how citizens interacted with government agencies.
The first feature it pushed allowed people to pay their parking tickets online. At the time the idea caught the attention of many investors with deep pockets. GovWorks became a media darling, the poster child of a technology boom that promised to lead to unprecedented prosperity.
GovWorks's story was so intriguing that a documentary crew followed Mr Herman and fellow co-founder Kaleil Isaza Tuzman around, as investor money flowed in and the company started hiring at a breakneck pace.
But with smartphones non-existent at the time and many Americans still without high-speed internet, GovWorks's business model fell far short of expectations and the company was among the many dotcom start-ups of the era to crash.
But Mr Herman – who has since founded Meta Carbon, a company that specialises in using technology for offsetting emissions – does not necessarily see parallels to current concerns over an AI bubble.
He said that although the dotcom crash caused a lot of headaches for investors and entrepreneurs, the overall arc of the internet prevailed.
"I think the internet changed the world probably more than all the hype originally," Mr Herman said.
Unlike the dotcom crash, in which some companies' ideas were ahead of the technology used by consumers, AI is creating solutions and getting people results, he said.
Mr Herman added, however, that some parts of history are being repeated.
"I recognise that the $5 trillion valuation of a company like Nvidia indicates an expectation of increased revenue and usage that seems almost inconceivable without cooking the planet," he said, reflecting on the electricity needed for AI data centres.
"The hype around AI is justified but the valuations are not."
Mr Herman said that in the late 1990s, GovWorks raised $60 million from investors on "almost nothing but an idea", and that ultimately became a problem when it came time to introduce a product in a new internet era.
GovWorks was hardly alone: there were countless website-based companies that were largely conceptual and attracted an influx of cash. Some even went on to initial public offerings, leading to the popularity of the phrase at the time, "Concept IPO".
Yet as Mr Herman points out, the overall technology infrastructure and AI tools offered are far more mature than what was available in the dotcom era.
Tim Bajarin, who has been a technology analyst and consultant for almost five decades, echoed Mr Herman's assessment of the current AI bubble fears.
"What we're heading into actually is a correction, not a collapse," Mr Bajarin said. He added it was not irrational for some investors to be concerned, but the fact that the technology behind AI is "richer, deeper and more focused" will help to soften any sort of disastrous predictions.
"We're more likely to see a normalisation period where investments will become more disciplined," Mr Bajarin said.
Some have pointed to the downfall of hyped companies such as Builder.AI, which once promised that its AI tools would make mobile app development as “easy as buying a pizza", as proof of a looming crash. But Mr Bajarin said those recent failures offer learning and are the exception to the rule.
"If you're using AI as window dressing, as opposed to solving a real problem, that's where I think you're going to see the separation of the wheat from the chaff," he said, adding that AI tools offered by OpenAI, Microsoft and Google are solving real-world problems for consumers and businesses.
"People just building something to call it AI doesn't bring the value to the proposition." Unlike the dotcom bubble, he said, many people have already got used to AI tools.
"It's not just for businesses, although they're taking great advantage of it. We're seeing people actually creating value using AI," Mr Bajarin said.
On Wednesday, semiconductor giant Nvidia, which has largely ridden the AI wave to unprecedented profits, helped to soothe worries of a bubble with its fourth-quarter results beating Wall Street expectations.
But even after Nvidia's positive results, markets still proved to be unpredictable on Thursday, partly prompted by continuing AI bubble fears and speculation that the Federal Reserve would not be cutting rates in December.
Comments made earlier in the week from Google's chief executive Sundar Pichai are also continuing to reverberate. Mr Pichai told BBC News that “there are elements of irrationality” with AI investments.
Regardless, Mr Herman, who also serves as chief technology officer for a startup, EcoClaim, said that even if there is a market correction with AI investments, the fundamentals surrounding the technology are strong.
This time, with the dotcom boom and bust well studied, he said the insights gained have not been forgotten.
"We have some wisdom that allows us to know where to focus and what to do," Mr Herman said, adding that he uses AI tools and apps in just about everything he does for work.
"When you apply wisdom to a lot of AI automated tasks, you can get a lot done, so harness it."
