Icons of Google's AI (Artificial Intelligence) app BardAI (or ChatBot) (C-L), OpenAI's app ChatGPT (C-R) and other AI apps on a smartphone screen in Oslo on July 12.OLIVIER MORIN/AFP/Getty Images
Awe-inspiring? Anxiety-inducing? Apocalypse-inciting? Eight months since ChatGPT’s debut sparked an artificial intelligence craze (and a Canadian privacy probe), society still can’t put its finger on what AI really implies. A recent survey found that 63 per cent of Canadian respondents worry about AI – well above the 52-per-cent global average. Yet another camp sees it as an express ticket to riches – the next dot-com boom. Which is it? You needn’t choose. AI holds opportunity and risk. Just keep in mind that both sides tend to overstate its ramifications.
Wildly off-base misperceptions partly underpin today’s exuberance and terror, with the term “AI” drumming up visions of sci-fi robots, good and evil. The reality is less viscerally exciting but just as interesting: AI attempts to mimic human consciousness and uses machine learning algorithms to generate text, images, video, computer code or other outputs based on mountains of data. ChatGPT, just one application among many, is a form of generative AI known as a large language model, capable of generating almost-human text via simple prompts. It seems intelligent because it answers questions, but in truth it is merely regurgitating prior programming.
Threats abound regardless. Among them: “deepfakes,” AI-altered voices, photos or videos; identity theft scams using AI-cloned voices and stolen personal information; cybersecurity concerns; and an alleged AI investment “bubble.” The OECD claims that AI puts 27 per cent of its member countries’ jobs at risk of elimination. In the U.S., one cybersecurity official warned that AI is a potential human “extinction event.” Regulations such as Canada’s slow-moving Artificial Intelligence and Data Act do little to quell such concerns. Conversely, the AI faithful say it will “save the world,” suggesting a utopian future and massive early-investor gains.
All camps miss the mark. Start with AI’s supposed economic and market magic. Sure, AI hype helped drive global tech stocks’ 38.8-per-cent gains so far this year. Semiconductors – crucial for AI development – led the rise, soaring 67 per cent. But tech’s 2023 rally is mostly about quality large-cap growth rebounding from 2022′s outsized slide.
‘I hope I’m wrong’: Why some experts see doom in AI
Consider: There has been no global or Canadian recession – yet. Value stocks didn’t get pounded to set up a big rebound, as in normal bear markets. Yet many still fear recession ahead while sweating the current slow growth. Therefore, markets pay a premium for firms that are able to increase revenue assuredly in a torpid economy. Those are real growth stocks – and they are quite rare, mostly huge ones.
Hence, the largest growth stocks lead now. AI is part of that – largely within semiconductors – but it isn’t alone. And the largest tech stocks are investing the most in AI. But we aren’t in the real game-changer phase of AI now. Having polled hundreds of firms, I can say its practical uses today are overall rather mundane, mostly things like automating repetitive tasks, back office work and marketing fluff.
Note, AI isn’t altogether new. Canada’s “godfathers of AI” began researching artificial intelligence in the 1980s! Tech startups pursuing AI attracted venture capital for years before ChatGPT came along. Big Tech firms used profitable divisions to subsidize AI research and development, providing an advantage over most tiny startups, such as Montreal’s now-defunct Element AI, because the computing power needed to train these systems is massive – and massively expensive. Hence, big players in chips, software, data analytics and search dominate AI now.
Identifying far-flung, long-term winners is futile. If I could, I would. But even after Silicon Valley’s gruelling startup shakeout, excess enthusiasm over tiny AI pure plays remains. Buying among them now is buying hype high. With oodles of ChatGPT clones already available, which startup will build the profit margins to justify premium valuations? You can’t know.
What about AI bubble concerns? Some headlines tout an AI investment deluge, inflating a bubble soon to burst. Bank of America claims AI is in a “baby bubble,” noting hot returns, yet not “white-hot” as in past full-blown bubbles, undercutting worries. AI’s social impacts? Whether on jobs, privacy, security or something else entirely, they are entirely unknowable. Good things could happen. Or bad. Or probably both! Or they could amount to little, like widely hyped self-driving cars or last decade’s at-home 3D printer hype. Remember that? Do you have a 3D printer at home?
Societal impacts will evolve over many years, while markets move on trends affecting corporate earnings approximately three to 30 months out – as they always do – not further. Current AI scams present some immediate risk, yet the reality of these scams isn’t new or unique to AI. There have always been rats. Beyond flashy window dressing, these scams are structurally similar.
So be clear-headed about AI – don’t succumb to fear or hype when investing. While some AI exposure may be beneficial, it shouldn’t be the swing factor to own or avoid a stock or sector. Seek high-quality growth – then decide if AI partly fuels it or not.
Ken Fisher is the founder, executive chairman and co-chief investment officer of Fisher Investments.
Be smart with your money. Get the latest investing insights delivered right to your inbox three times a week, with the Globe Investor newsletter. Sign up today.