They called it a ‘genius’ AI. It turned out to be two men simply writing things down

The story of a billion-dollar tech company that pretended it had a whizzy note-taking AI when it was actually two very normal humans says everything about Silicon Valley’s fake-it-until-you make-it culture, says James Moore – and explains why we are all in danger

There is a story going around that I think sums up this whole “AI bubble” neatly.

It is about a startup that pitched itself as the developer of a “genius" AI that could write up meeting notes from video calls and send them to you afterwards.

That’s what they said it was. But what it actually was, at least in its early days, amounted to two blokes who would join that meeting on mute, take notes by hand themselves while sitting silently and send the “AI-generated” notes 10 minutes later.

“We told our customers that there’s an ‘AI that’ll join the meeting’. In reality, it was just me and my co-founder calling into the meeting, sitting there silently and taking notes by hand,” Fireflies AI co-founder Sam Udotong wrote on LinkedIn this week.

“After taking notes on 100+ meetings (and falling asleep in many), we were finally able to pay $750/month rent for a tiny SF living room. That was the point when we said, ‘Let’s stop and automate everything.’” That company is now worth $1bn (£760m) based on its latest fundraising.

There are shades here of Kimi, Steven Soderbergh’s techno-thriller, whose human protagonist has the job of listening in to users of the eponymous app to improve its understanding of our voices.

Whether this really is a sign of the slow-burning pop of the AI bubble remains to be seen, in the week that Japan’s SoftBank dumped a bunch of Nvidia shares, whose chips are what the AI edifice is built upon.

SoftBank, a tech investment outfit, will cycle the money into, you guessed it, more AI, the funding requirements for which almost make even Elon Musk’s $1tn Tesla salary look modest, with little enough (so far) in the way of returns to show for it.

What is clear is that Silicon Valley is built upon fear and greed. The greed explains itself, but the fear is of missing out on the next big thing. Investors feel it every time they see a PA driving a Porsche, bought after cashing in the Google stock options they got when they joined the company before anyone had heard of it.

This can lead very rich and ostensibly smart people into deep black holes. An obvious example would be those who poured millions into Theranos, with its (false) claims of revolutionary diagnostic blood test tech. It can lead to crazy valuations until someone somewhere shouts, “The emperor’s got no clothes!”.

Perhaps that’s why the FTSE 100 is steaming towards an all-time high, near the 10,000 point mark, with its corps of dull but reliable dividend payers with proven (if not always savoury – looking at you, Big Oil and Tobacco) products. The market has long been considered defensive by those who worry about bubbles popping or the global situation turning to hell.

Tech may be the belle of the ball, but there will always be a place for, say, retail banks that offer savers deposits and provide mortgages, or a Tesco that does groceries really well, or an AstraZeneca that has a portfolio of proven medicines people rely on.

In the meantime, I suspect that Fireflies CEO Krish Ramineni is going to have to spend the next few days on the old-fashioned phone to convince investors that the company isn’t another bust. They can prove quite sensitive about this sort of story.

We still don’t know whether AI is the big win it’s cracked up to be, the dawn of William Gibson’s future shock tech thriller Neuromancer, with its all-powerful super-AIs, or the latest example of over-frothy techno-hype.

This sort of tale is, however, grist to the sceptics’ mill.