The boffins at Stanford University reckon there has been whopping $2 trillion of investment in AI over the last five years alone.
Subscribe now for unlimited access.
or signup to continue reading
With numbers that big, it's got many wondering whether these investments are sensible or whether we are in an AI bubble.
There are lots of opinions being thrown around, but they all have one thing in common: we really have no idea.
Many are pointing to the huge benefits and transformations that AI could have for our economies and societies. Others are sceptical.
These are both interesting perspectives, but the real question is whether those making the big investments can make any money out of them.
This is as much about economics as it is about technology.
If this new tech has a big impact and barriers to entry are high (meaning that it's hard for new firms to come in and start competing away profits at different levels of the AI stack), then these firms will probably make a lot of money (a.k.a. no bubble).
But if barriers to entry are low and the tech turns out to deliver underwhelming results, then some of those investments might not be looking so good.
Truth is, we don't know the answers to these questions and are only guessing whether there's a bubble or not. So perhaps we should be asking an easier question: do we care if there's a bubble?
Here's a thought experiment: suppose there is a bubble and suppose the world's investors and tech bros suddenly change their mind about AI and the bubble bursts. What would happen?
For Millennials like me, the idea of economic turmoil fills us with dread.
We entered the workforce during the global financial crisis, worked through the European debt crisis, had the COVID-19 recession and lockdowns with young kids and are still reeling from the cost-of-living crisis that followed.
Put simply: all the economic recessions we've had during our adulthoods have been catastrophic.
But this is not always the case. Some recessions are actually quite boring.
The recession in 2001 - the so-called Dot-Com crash - is a case in point. It's also probably the best analogy for what an AI-bubble-pop-recession would look like.
The dot-com crash involved what Alan Greenspan, former chair of the Fed, called "irrational exuberance" - basically when investors and markets get way too excited about something.
The thing they got excited about in the '90s was the internet. Any start-up that had ".com" in their name had money thrown at them faster than they could say World Wide Web.
Venture capital investment rose sixfold from 1995 to its peak in 2000. The stock market followed suit. The NASDAQ soared from 1000 points to more than 5000 points over the same period.
Greenspan gave his "irrational exuberance" warning in 1996. But nobody cared. That was, until, the bubble burst in 2001.
The result was a big crash in the stock market. The NASDAQ plummeted from 5000 points back down to 1000.
Countless companies went under - from Pets.com (that sold petsupplies), Webvan (that delivered groceries) and Boo.com (online apparel) through to big communications companies like WorldCom, NortPoint Communications and Global Crossing.
Cisco lost a whopping 80 per cent of its stock value. Others got bought out. The US economy went into recession.
So, how bad was it?
By modern standards, it was pretty mild. The recession lasted only eight months and wasn't particularly severe in terms of job losses.
Employment fell about 2 per cent. Employment fell three times that during the global financial crisis and seven times that during COVID-19.
Economists call it a "standard recession" because there was nothing particularly unique or scary about it.
A bunch of private-sector investors made a bet. The bet didn't pay off.
They lost their money. The Fed cut interest rates to get the economy back on track and, within eight months, it was.
This is in stark contrast to the global financial crisis which involved an unknown quantity of toxic assets sprinkled across an unknown number of banks and financial institutions across the global financial system.
Risk premia skyrocketed. The world's biggest and oldest financial institutions collapsed. Countries were toppled. The European Union was almost destroyed.
COVID-19 was just as crazy. In the early months, we had no idea how this virus worked or whether this was the end of humanity.
READ MORE ADAM TRIGGS:
Economically, it caused demand to collapse, then caused global trade and supply chains to freeze, only then to see demand skyrocket and an inflation crisis ensue.
Compared to all that, the dot-com crash was a breeze.
And this brings us back to AI. If you read the above story about irrational exuberance and thought "this sounds familiar", then there's good news: if the story ends the same way as the dot-com crash, it probably won't be that bad.
There'll be two main victims. The first will be the investors who lose some their money and those connected to it. The second will be president Donald Trump whose terrible economic policies will no longer be hidden by unprecedented levels of tech investment.
As Warren Buffet says, "it's only after the tide goes out that we learn who's been swimming naked". AI bubble or not, now would be a prudent time for America to put some clothes on.
- Adam Triggs is a partner at the economics advisory firm, Mandala, and a visiting fellow at the ANU Crawford School and a non-resident fellow at the Brookings Institution.
