The processing power behind AI
Nvidia makes the microchips that power artificial-intelligence applications, giving it a bright future
For growth investors, a positive theme this year has been artificial intelligence (AI): computers thinking and solving problems just like people do, allowing them to play bigger roles in everyday life. Many commentators see this as a transformative and profitable “mega-trend” in society, so stock prices of anything involved in AI have been soaring.
The sector has gained nearly 17% in 2023, more than double US stocks overall, as measured by the near 70 AI-related stocks in the ROBO Global AI Index. A catalyst is ChatGPT – a new app that accepts requests in everyday language and responds helpfully in a conversational, “human” style. It draws us closer to the internet because it uses the greatest human interaction device that we have: our ordinary language. This brought AI to a broader audience beyond researchers and has inspired new and ambitious ways of thinking about future possibilities.
Propping up the market
You could say ChatGPT has been keeping stocks afloat this year. Strip out the gains from AI and the market looks sickly – Société Générale estimates that without AI, America’s S&P 500 index wouldn’t be up 8% this year – it would have fallen by 2%, a 10% reversal. So given that AI is one of the very few positive opportunities this year, should investors be drawn in? One thing we can say for sure that dampens some of the hype is that we’re nowhere near developing “sentient” – “feeling” or “thinking” – computers. But what we are getting better at developing is the increasingly powerful microprocessors that can keep absorbing and learning from the vast amounts of information fed to them. The chips use this to make high-speed connections and decisions across that database which gives the impression the machine is thinking.
A simple everyday example is a computer that’s absorbed every movie and TV series ever made, and so can “decide” simultaneously what millions might enjoy next based on what each of them has watched in the past – that’s a lot of “thinking”. Or consider a computer that has millions of hospital records and can “see” new and previously unknown patterns in the data to predict potential serious illnesses in otherwise healthy people all over the world.
Fast chips
Chips, then, are key. The uses for AI are limited only by our imagination – there will be good apps and gimmicks. But either way, to do anything, you need the fastest microprocessors that make AI happen. Their makers have much to gain.
It is no surprise, therefore, to see Nvidia (Nasdaq: NVDA), one of the world’s most innovative chip designers, leading the AI sector higher with a gain of nearly 100% this year – the company’s market value is now $716bn. It has been a long-time favourite and holders should sit tight. It is hard to escape the hype and short-term speculators will be behind some of the surge in the stock price. But to sell creates the dilemma of how to time getting back in – get it wrong and you might find yourself locked out of an attractive long-term growth story.
New buyers also face a dilemma. The stock is highly valued given the hype over AI and the uncertain economic backdrop. On the other hand, this well-run outfit is expected to deliver, and the superior pricing of its products is holding up well. AI is not a fad but will drive efficiencies and profitability across diverse industries for years to come. To compete, firms must invest. The more adventurous readers who appreciate the long-term demand that exists for evergreater microprocessing power, which few offer alongside Nvidia, and who can overlook near-term market volatility, can read more below.