How Microsoft and NVIDIA Made the Right Bets to Overtake Apple

Last month, AI chip giant Nvidia briefly became the world's richest company, overtaking Microsoft, which has since overtaken Apple.

When this news was mentioned onstage at a tech industry event I attended in Copenhagen, the audience burst into spontaneous applause.

As I write this, falling share prices have seen Nvidia's total market capitalisation fall to $3 trillion (£2.4 trillion), putting it back in second place compared to Microsoft's $3.4 trillion.

Two factors have propelled America's two tech giants to such astonishing heights: AI and visionary thinking.

Microsoft began investing in OpenAI, developer of the popular AI chatbot ChatGPT, in 2019. Meanwhile, Nvidia president Jensen Huang had been pushing his company towards AI chip development years before generative AI exploded in popularity.

Both companies are making long-term bets on the current AI boom, and so far that bet has paid off, putting former leader Apple in hot pursuit. But how long will this bet last?

This year's London Tech Week, the UK tech world's annual event, might as well have been called London AI Week, with the word AI plastered on every booth and uttered in every speech.

I bumped into Anne Boden, founder of Starling Bank, a major game changer in the fintech industry, and she was buzzing with excitement.

“I thought I knew who the winners and losers were. [in tech]“But with AI, you’re rolling the dice again,” she told me.

She sees the AI ​​revolution changing the face of the tech industry and is keen to dive back into that field.

That same week, I also attended Founders Forum, an annual gathering of about 250 top entrepreneurs and investors. That means a lot of money. It's a closely guarded event, but I don't think it matters much if I say that the majority of the conversation there was about AI, too.

Life goes by really fast.

“Given the rapid rise in tech company valuations, any missteps in the future could cause big volatility in share prices,” warned Susanna Streeter, head of funds and markets at investment firm Hargreaves Lansdown.

“Just like the dot-com bubble, there is a danger that excess enthusiasm will lead to disappointment.”

It's reasonable to assume that in 2023, anything with the AI ​​acronym attached will see a flood of investment capital into all things AI-related, certainly opening up a tier of lucrative capital.

My friend Saurabh Dayal, who is based in Scotland, has been identifying AI projects that his investment firm could potentially collaborate on.

He said he quickly grew tired of the misleading sales pitch.

“I keep saying, '…but that's not AI,'” he says.

Investors and customers alike finally seem to be coming to terms with the term AI and are becoming more cautious as a result.

Additionally, there is a growing awareness that current generative AI products are not living up to expectations: inaccurate information, misinformation, displays of bias, piracy, and just plain weird content.

Additionally, early AI-enabled physical devices such as the Rabbit R1 and Humane Pin have received poor reviews.

“The market for generative AI is slowly maturing. Early experiments have shown great promise, but real-world applications have yielded too many unexpected results,” said Chris Weston, chief digital and information officer at technology services company Jummah.

“For businesses, the trust and confidence that customers have in their services is a huge value that is tied to trust. Deploying uncontrollable chatbots is a step too far for many businesses at this point.”

Technology analyst Paolo Pescatore agrees that AI companies are under pressure to deliver on their promises: “The moment one of the giants fails to demonstrate meaningful growth from AI, the bubble will burst,” he said.

But he doesn't think that will happen anytime soon.

“Everyone is still fighting for position and every company is committing their strategy to AI,” he added.

“All players are stepping up activity, increasing spending and claiming early success.”

There's another reason the AI ​​bubble will burst, and it has nothing to do with the quality of the products or their market value: it's whether the planet itself can tolerate it.

A study published last year predicted that if the AI ​​industry continues to grow at its current rate, by 2027 it could consume as much energy as a country the size of the Netherlands.

I interviewed Professor Kate Crawford from the University of Southern California for the BBC Tech Life podcast, and she said she keeps sleepy at night worrying about the amount of electricity, energy and water needed to run AI.

Dr Sasha Luccioni of machine learning company Hugging Face is also concerned.

“Right now we simply don't have enough renewable energy to power AI; most of our energy comes from oil and gas,” she says.

It is hoped that the technology can be used to identify sustainability solutions, for example the secrets of nuclear fusion, how the sun gets its energy. But that is not yet the case, and in the meantime, “AI systems are putting even greater strain on an already heavily strained energy grid,” adds Dr Luccioni.

With so much uncertainty, few would bet on not seeing another round of restructuring among the world's richest companies, but right now Apple is struggling to catch up with Microsoft and Nvidia in the AI ​​race.