Artificial intelligence is underappreciated rather than overhyped, finance expert Ulrike Hoffmann-Burchardi (UBS) argues in her engaging DLD Munich 2025 presentation about the business potential of AI.
“AI, unlike any other technology before it, is self-learning”, she points out. “So the current Large Language Models are training the next generation of Large Language Models.”
This capability creates progress “at a much faster rate than anything we have seen before”, Hoffmann-Burchardi says. “In fact, we believe that AI is the most profound innovation in human history.”
She disputes the common perception that capital expenditures (CapEX) are already too high. “CapEx is only going to go up”, the highly regarded UBS analyst predicts, reaching about $400 billion by 2027. “We will spend more on data centers than we have spent on fiber cable optics in the late ’90s”, Hoffmann-Burchardi says.
“This sounds massive”, she admits – but improving the current AI models will require more computing power, and the payoff promises to justify the expenses. “We think AI will be more than worth the current investments”, Hoffmann-Burchardi notes, quoting two examples.
Walmart used generative AI to update its product catalog. “The ROI was 100x.” Amazon saved $260 million upgrading software applications. “These data points that I cited will be the weakest that we’ll ever hear about”, Hoffmann-Burchardi emphasizes.
Watch the video for further insights into the evolving landscape of AI investment, the disruption of traditional software companies, and how AI is set to reshape the global economy at an exponential pace.
