If this is an AI bubble, then heed the lessons of the dotcom bubble

 Are we witnessing an AI bubble? If this is an AI bubble, then remember that the dotcom bubble was followed by an Internet revolution.

It's a shame about the Internet. Do you remember the Internet? There was a time when hype suggested it would be the next big thing. But of course, as we all know, it fizzled out, and nowadays, no one uses the Internet. I say this with certainty because I read it. In December 2000 the Daily Mail ran an article saying the Internet may be just a fad. And since newspapers never lie, the Internet must have fizzled out and gone the way of hula hoops and South Sea investments. 

 Forward the clock 24 years, and the Daily Mail has run a similar piece, this time about AI. "The artificial intelligence experts who believe the AI boom could fizzle or even be a new dotcom crash: 'We are starting to see signs it might be a dud' headlined the newspaper. 

So, are we seeing an AI bubble like the dotcom bubble of the 1990s?  Well, maybe we are. Indeed, the chances of some kind of crash in AI-related shares are not remote. Things happen that way. Stock markets and the media get overexcited and then crash; all those hopes are left to crash against the rocks of reality like ships on the sea of dreams.

 But was the Internet a fad? I have some breaking news. The Internet is still with us. Today, most of the world's largest companies have a business model that could not exist without the Internet. Consider the combined market cap of these companies. Apple, Alphabet, Meta, Nvidia and Amazon have a combined market cap of around $9 trillion. (I exclude Microsoft from the list.) The value of Internet-reliant companies is at a level that makes the wildest dreams of dotcom entrepreneurs seem surprisingly mild.

Because whilst dotcoms did crash, and whilst a popular narrative emerged in the aftermath that the Internet miracle had fizzled out, leaving nothing more than a few wiser heads, the truth is that in the years that followed the crash, the Internet changed the world. Broadband helped. Maybe if we still had to dial up to the Internet via our telephone listening to a discordant symphony of zeros and ones, the Internet may have stalled. Smartphones had something to do with the revolution that followed. Maybe if we still had to sit at our desks glued to our computers to use the Internet, its mark upon the world would not have been so great. But that is no excuse for the Internet denialism that followed the crash — the Internet always was incredible technology waiting for a means to beam it into daily lives.  On its own, the Internet was not enough to change the world, but convergence between the Internet, broadband, smartphones and then technology for creating super-fast wireless connectivity created a permanent change. 

 The AI rollercoaster 

Where will the AI rollercoaster go next? Well, it may go down, leaving investors' hearts in their mouths and creating smug 'told you so' comments by those who predicted a crash.

But then the rollercoaster will go shooting up just like the Internet did. 

It is when we see convergence that things will really get interesting. What form will that convergence take? Personally, I believe we will see the emergence of AI personal assistants, always with us like smartphones are today, but with an even greater impact on our lives. These AI personal assistants may work in conjunction with an augmented reality device. In time (but not for a couple of decades), we may see a convergence between AI, augmented reality and brain computer interfaces, a convergence which will see our AI companion know us better than we know ourselves. (For that matter, AI may not even need brain-computer interfaces for this to happen.) 

Of course, AI will also be applied in R&D, helping develop new disease therapies and helping us learn how to extend lives. It will be a key part of the renewables revolution working out how to optimise the match between the supply of energy, when conditions for renewables are ideal with demand. AI will help in the workplace too, taking on more and more dull, routine tasks but also supporting us in more creative ways. Maybe, most importantly of all, AI will be employed to design AI and help create more powerful computers, increasing the power of AI.

It is not clear where the AI rollercoaster is heading. Is it on its way to the sunlit uplands or somewhere more pernicious? 

Maybe we are heading for an AI singularity, a point in which AI advances so rapidly no one can keep up or understand it.

 Why AI denialism is dangerous

 "So what," you might ask, "if we get a bit of AI denialism? It doesn't really matter?" But it does matter. To argue that AI is no big deal, that it has been hyped beyond reason, is dangerous in much the same way climate change denialism is dangerous. AI may transform humanity's existence for the better. It may represent an existential threat to humanity. Only by seeing AI for what it is, the risks it poses and the opportunities it brings can we start to figure out how we can ensure AI's impact upon humanity is as benign as possible. If, instead, we have to stop in our tracks and, in the face of reason, defend the position that technologies such as AI are indeed creating extraordinary change, we risk taking our eye off the ball and instead focus on the wrong thing. AI denialism is an unnecessary distraction. 

 What we can do 

Whether the outcome will be good or bad is unclear, but we can play a role in helping to push it along a more positive track.

 Regulators must play a role in the development of AI. But regulators will always struggle to keep up with business innovation. For that reason, it is imperative that companies exercise considerable responsibility.

 AI needs ESG 

This takes us to ESG. Companies must apply a kind of ESG mindset, see the bigger picture, the societal impact their actions might have, and apply smart governance. 

 It is odd. Out-and-out fans of the free market don't like ESG and think businesses should be free to focus on profits before anything else. Those same people don't like the government interfering with government or bureaucracy; they say private enterprise is the most effective way of creating wealth. If that is true, and AI represents both opportunity and danger for humanity, is it not best for private enterprises to focus their creativity and ingenuity on seeing the bigger AI picture and exercising responsibility? 

The malaise of the backlash

But there was one enduring legacy of the dotcom crash. Before the crash, the UK was grasping the internet possibility and the City backed it. Baltimore Technologies, for example,  (based in Ireland but listed on the London Stock Exchange) joined the FTSE 100. But the crash left a bitter taste among London traders and financiers, weakening the UK economy and leaving it in the corporate slow lane. Its tech stars sought solace in the arms of the US, with DeepMind, for example owned by Alphabet and ARM listed on the NASDAQ (and partially owned by SoftBank.) The British sense of cynicism has created wonderful comedy, but it has sucked dynamism out of the tech sector, created a stock market reliant on low-growth companies like banks and oil firms and may explain much of the economy's current weakness. It was as if the dotcom crash was like manna for heaven for British sceptics whose creed of negativity has left a wound in the UK that is showing little sign of healing.