Martin from Globalt argues that the current surge in AI development is not a bubble but a reflection of genuine early-stage adoption driven by the technology’s real usefulness and significant computational demands. He highlights challenges such as geopolitical risks, managing expectations without immediate productivity data, and compares AI’s rapid adoption to the early internet, emphasizing that ongoing investments and future business integration will ultimately validate AI’s long-term potential.
In the discussion, Martin from Globalt addresses the question of whether the current surge in AI development represents a bubble. He argues that it is not a bubble, emphasizing that the driving force behind the growth is the genuine usefulness of AI models. According to him, we are still at the very early stages of AI adoption, with companies experiencing varying levels of success in implementation. The enormous computational power required to run these models is a significant factor, and ongoing deals in the industry reflect efforts to provide this compute capacity to businesses that need it.
Martin acknowledges that concrete data showing productivity gains and revenue increases from AI is not yet fully available, which fuels skepticism. Much of the current investment is based on the expectation that AI will solve complex problems and become highly productive in the future. While consumers can already see improvements in AI’s ability to answer questions, the real impact will come from sophisticated business applications involving automation and intelligent agents, which will demand even more computational resources and connectivity.
Geopolitical risks, particularly related to supply chains for rare earth metals and magnets essential for AI hardware, are significant concerns. China controls a large majority of these materials, giving it considerable leverage that could disrupt supply if tensions escalate. Both China and the United States are aware of this power dynamic, and any conflict could have serious consequences for the AI industry’s hardware supply. However, Martin believes that the market’s recognition of AI’s vast potential will motivate efforts to prevent such disruptions.
The conversation also touches on the challenge of managing client expectations in an environment where tangible proof of AI’s economic impact is still emerging. Martin compares the current AI wave to the early days of the internet, when infrastructure investments outpaced immediate returns. He notes that while concepts can be exciting, real value is realized only when technologies are widely adopted and integrated into everyday business operations. The pace of AI adoption, however, is much faster than previous technological revolutions.
Finally, Martin highlights that recent deals and investments in AI are concrete evidence of serious financial commitment from a diverse group of sophisticated and well-funded players. As companies report their earnings in the coming quarters, the market will be closely watching for signs that AI is delivering measurable productivity and revenue growth. This ongoing scrutiny will help validate the long-term potential of AI beyond the initial excitement and speculation.