Artificially hyped

Hey there, time traveller!
This article was published 19/08/2023 (792 days ago), so information in it may no longer be current.

Artificial intelligence makes sliced bread seem like a vastly distant second as the next best innovation for humankind.

While artificial intelligence — AI for short — had long been more science fiction than real science, with the recent unveiling of generative AI tools such as ChatGPT and Microsoft’s new AI-driven Bing search engine, our imagination has been captured by the potential of intelligent machines.

For investors, AI has also become a source of speculation and for good reason.

A recent research report by PWC forecasts AI could contribute nearly US$16 trillion to the global economy by 2030.

AI also has the potential to make us better investors, though those upsides are by no means slam dunks, at least for now, says Massachusetts-based Dave Nadig, a financial futurist and writer with VettaFi, an investment analysis and news platform.

“My job is to figure out how things will change in the next 10 to 15 years for investing and right now, we’re in a hype cycle for generative AI, which is likely to have much less impact on how we invest than we likely expect.”

Nadig notes the current technology now widely available through ChatGPT, which people can sign up to use for free, is a tool that certainly can tell you how to build the ideal balanced portfolio or find you the top Canadian equity mutual fund.

But does that actually mean it is providing a correct answer? The answer is no. In fact, it’s likely not going to offer you any advantage over a typical Google search, Nadig says.

“When you say, ‘Pick me a good value fund,’ literally all ChatGPT is doing is taking the phrase ‘good value fund’ and looking in its database for the best words that might come next.”

That is what the software is designed to do, use language models to pore over vast amounts of information, determining statistically what is likely the next best word to come after the previous one. This works exceedingly well — most of the time. For example, it can write the Declaration of Independence word-for-word because it has seen it so many times. It’s trained on the data pertaining to that well-known document.

In the investment industry, generative AI has already been in use for some time, Nadig says, adding that mathematician and hedge fund manager Jim Simons has used proprietary AI software which has helped provide annualized returns over several years averaging nearly 40 per cent.

Yet its benefits are largely dependent on the information AI can access and how it is being trained. The machine learning component, to use the data, and that last part, the data, is critical, Nadig adds.

“The reason hedge funds can outperform me as an investor has nothing to do with access to better tech and everything to do with the fact that they pay for better data and know what to do with it.”

Still, AI will likely become an important tool for smaller investors in the next few years as the technology becomes more refined and widely offered by brokerages, says a veteran fund manager and founder of a firm providing AI-driven software for environmental, social and governance (ESG) investing.

“I do see AI helping to level out the playing field between large institutional investors and retail investors in the future,” says Toronto-based Douglas Chow, chief executive officer (CEO) of PortageBay.

He notes that AI is already decreasing the cost of analysis by “about 100-fold” because it allows investors to get pertinent data from a stack of documents that are hundreds of pages long within seconds.

“That ability can potentially democratize analysis.”

Besides its potential as a useful tool to analyze investments, AI also presents an investment opportunity, though this too may be somewhat overinflated at the moment, Nadig says.

“The entire AI market is dominated by giant tech companies and very small private pure-play companies like OpenAI and Stable Diffusion.”

Most investors likely already own Microsoft, NVIDIA, Alphabet (Google), which have made big investments in AI, through mutual funds and exchange-traded funds.

These companies’ recent uptick in share price has been driven by their endeavours in AI, but most of that price appreciation is a result of investor hype as opposed to AI’s actual impact on revenues. At the moment, the technology still represents a sliver of their overall businesses, Nadig says.

In contrast, those pure-play, private firms developing AI technologies, which could grow wildly in value, or go bust, are largely inaccessible to retail investors because they do not list on any stock market.

But AI is likely to benefit some non-technology firms by helping them automate low-skill jobs.

That includes fast-food chains, which are already experimenting with AI chatbots at the drive-thru.

“Those low-level customer service jobs will be the ones replaced the very fastest by AI,” Nadig says.

He further points to a McKinsey & Company report predicting the technology could displace 12 million jobs in the next six years.

As AI becomes more powerful, it may threaten more skilled occupations too, a concern shared by striking screenwriters in Hollywood, Nadig adds.

Yet those fears actually coming true are likely a few years away and AI that truly mimics human intelligence is probably many more years down the road.

That said, things can change quickly as we’ve already seen — noting generative AI wasn’t available to the public even just a year ago, Nadig says.

“The big challenge in AI development is it’s reached a point where the changes and growth rates are no longer incremental, they’re exponential.”