"Superintelligence" Is on the Horizon: Deutsche Bank Identifies Three Major Challenges for Investors

Deutsche Bank predicts that the best artificial intelligence models could reach the stage of “superintelligence” by 2030, surpassing human experts in all fields. This will require changes to investment rules: market participants will have to assess companies’ prospects differently, rethink their approach to risk, and stop automatically buying stocks that have fallen in price, according to analyst Adrian Cox, as quoted by Yahoo Finance.

According to Deutsche Bank’s scenario, AI models with capabilities surpassing those of humans will be available virtually everywhere: on every smartphone, tablet, or laptop, regardless of their processing power. “These will be cutting-edge models: using them will feel exactly like gathering the best human experts in one room to solve any problem a business faces,” wrote Adrian Cox.

This could lead to fundamental changes in business. For example, while Nvidia is currently training robotic arms to install graphics processors and cut through plastic ties, in a world of superintelligence, an entire factory could be retooled for new tasks overnight, according to a Deutsche Bank analyst.

"Superintelligence will mark a fundamental break from the previous regime, not just a faster version of what exists today," Yahoo Finance quotes Cox as saying. Instead of simply carrying out tasks assigned by humans, AI will independently define those tasks, the analyst believes.

Based on this, Deutsche Bank has identified three challenges that stock market participants may face following the emergence of super-intelligent AI.

First, learn to distinguish between companies that will benefit from the development of AI and those whose business models will be threatened. According to Cox, investors should classify companies as potential winners and losers from the technology, rather than focusing on average market indicators. Under normal circumstances, mistakes in asset selection can partially offset one another. But if the development of superintelligence leads to a sharp division of companies into two groups, this approach will no longer work. “The size of the position should correspond to the confidence in the accuracy of this classification,” the analyst noted.

Second, don’t confuse low volatility with low risk. Historical price fluctuations may not reflect the risks a company will face when a new technology emerges. Up to a certain point, the stocks of potential winners and losers may behave almost identically, but then their performance will diverge sharply.

As an example, Cox cited the educational software company Chegg. In May 2023, its stock lost half its value in a single day after management attributed the slowdown in new customer acquisition to the launch of ChatGPT.

Third, reconsider your position-averaging strategy. The traditional approach assumes that a drop in a stock’s price could be an opportunity to buy it at a lower price. “A falling price may be a sign that the business model is outdated, rather than a buying opportunity,” Cox warned.

This article was AI-translated and verified by a human editor