At Substantive Research, our job is to highlight the most valuable insights on key trending investment themes across the research market in 2023. This week, in part 1 of a 2-part series, we delve into the world of Generative AI, which has received its fair share of hype in recent months.
Part 1 features research that takes a 10,000-foot view of this new technology as analysts attempt to provide some perspective on all of the hype. What potential advantages does Generative AI offer to society, business, and the investment research and investment process? What are the risks? Part 2 will explore Generative AI as a potential investment theme.
Starting with the basics, we found Deutsche Bank’s comprehensive guide to AI useful. Their thematic research team discussed the history of AI, its various types, and how it can enhance business productivity, making it an excellent starting point.
There are high hopes that Generative AI will be a ‘game changer’ for productivity improvements. That was the focus of Liberum’s Joachim Klement in his excellent ‘’Thought for the Day’’ column, where he cited a trifecta of studies that presented evidence that Generative AI could really move the needle on productivity, and much more sustainably than the adoption of PCs and the internet in the 1980s and 1990s. As Klement points out, while both of these technological innovations were believed to be massive productivity enhancers, they were actually just a blip on the long-term productivity timeline. Klement suggests that Generative AI has the potential to surpass them and take productivity to a different level altogether.
From his own analyst’s seat, Klement also commented on how Generative AI had now liberated him of the many mundane tasks that are part of every research analyst’s daily grind. While some fact-checking is still required, the time saved on research was enormous, he said. Of course, Generative AI is just another stage in the evolution of technology in the research process. Check out this interesting post from Mike Mayhew at Integrity Research who has charted the evolution of technology in research over the past 20 years.
As for whether there might be some value in using Generative AI as an investing tool, well, that’s another question altogether, according to the research team at Man Group – a very quant-savvy hedge fund that is always at the forefront of leveraging technology to enhance returns. While the team acknowledges the benefits of incorporating AI into investment management, their own research found that combining signals generated by AI systems to make buy and sell recommendations resulted in a suboptimal outcome. In their opinion, investment management requires an appreciation of shades of grey because the number of parameters is often unknown and their strength is ephemeral. Man Group concludes that for the time being, human investors, although heavily supported by machines, are still needed in investment management. Your job is safe for now, Mr and Mrs Portfolio Manager.
In terms of the broader technology adoption spectrum, where new innovations are built on existing ones and trends emerge and fade, how should Generative AI be perceived? For insights, we deferred to Inferential Focus, whose work in the past decade has consistently focused on the development of the “digitally trained consumer” and the application of behavioral science.
While large language models (LLMs) are enchanting the digiterati, Inferential Focus (IF) says ChatGPT is just the latest in a long line of digital crazes. In recent times, we’ve had the metaverse, cryptocurrencies and NFTs. While Inferential Focus concedes that artificial intelligence does hold out many possibilities in the future, for now, the digital world – despite its momentary crazes – is losing its status as the “cool thing.”
According to Inferential Focus (IF), while businesses are increasingly adopting digital-first operations, digital technology is losing its cultural influence on behavior, thought processes, interpersonal communication, the popular arts, and societal values. IF argues that recent technological breakthroughs are reviving deep-seated fears of what technology might do to humans and how those breakthroughs can impact society’s future. This same idea is covered by Bank of America’s thematic research team who also focussed on some of these ethical, and also environmental concerns within their comprehensive AI primer, more of which we will highlight in Part 2.
In their follow-up piece, IF explored both the positives and negatives of Generative AI. To do that, they compared AlphaFold, a DeepMind technology that forecasts a protein’s 3D structure from its amino acid sequence, to ChatGPT, which made the erroneous claim that it took nine women one month to produce a baby. This was simply an example to illustrate that breakthroughs and errors have become commonplace in the realm of artificial intelligence. Be wary, was their conclusion.
We realise that what we have highlighted above is just a small sample of the vast reams of copy that has been written on Generative AI, but our aim here has been to provide some colour, and balance to this fascinating new innovation. If you wish to find out more about any of the research providers we have featured above, please reach out here.
Keep an eye out for Part 2 in the coming weeks, where we’ll delve into how, and where investors can put their money to work in nascent technology.
The full list of Substantive Research’s top trending themes in the past week is as follows:
- Oil
- US Banks
- Real Estate
- Commodities
- Artificial Intelligence
Too Much Research And Not Enough Time? We Can Help
With access to more than 250 research providers globally, across all asset classes, Substantive Research is a research aggregator with a difference. Our experienced team of curators filters and selects only the highest quality research, tracks the most relevant themes, then tags them in an intuitive manner. Another key difference between Substantive and other research aggregation platforms is that we pre-permission access to all of the research for you. For more visit our Research Discovery service below.
