Rethinking ROI in the age of AI
On 9th September 2025, the asset management division of J.P Morgan posed an ambitious question entitled ‘’Is AI already driving US growth?’’ on their portal. Examining data from the U.S. Bureau of Economic Analysis, it estimated that ‘’AI-related capital expenditures contributed 1.1% to GDP growth, outpacing the U.S. consumer as an engine of expansion.’’ AI has quickly evolved from a promising nascent technology to a dominant market theme attracting investments on a gigantic scale disrupting multiple sectors. The massive investments is prompting many to wonder out loud if the potential returns justifies the extent of investments. Is AI investment a bubble about to burst? Is AI an ROI positive wager from an investor point-of-view? Are comparisons with tulip mania or dot.com bubble appropriate? Big techs are pumping billions of dollars in GPUs, server systems and data centers while simultaneously asking these skeptical questions. Nobody is halting capital expenditure and risking being left behind. The stakes are too high for quite introspection on ROI in the age of AI.
The ROI on AI investments
Two contending ideas dominate the narrative on ROI. First one claims that AI disruption is real and sustainable but overhyped. Viewing from the prism of return-on-investment, they argue that AI company valuations are crazy making it unlikely to reap adequate returns in the long term. They point to dot com examples wherein amazon stock price declined by 94% between late 1999 and late 2001. After the bubble popped, it took 10 years for amazon to bounce back to the same levels. The prospect of portfolio being in the red for extended period is worrying investors even if valuations aren’t unreasonable.
The second explanation equally popular among investors is that while AI investments are appropriately priced for the potential impact, you never know which horse to bet on. At the height of dot com bubble, numerous companies like Pets.com raised millions of dollars despite losing money on each shipment. While the expectation of investors that consumers would eventually purchase pet supplies online proved prescient, Pets.com was already out of business by then exhausting all its resources. The investor community is worried that history may repeat itself even as AI companies raise billions and invest heavily. Big tech AI giants losing money on each prompt are reminiscent of hundreds of dot com companies that couldn’t stay in business when the funding dried up. Even though online ecommerce models proved resilient as a business, success stories like Google and amazon aren’t the norm but exceptions that prove the rule. The real answer to ROI question then is: which AI companies would be the last men standing to reap the benefits?
Apart from ROI from investors perspective, the technological capabilities of AI is an ROI worth pondering. The capacities can be broadly divided as follows:
1.Impossible tasks for humans
AI can solve certain problems that a human brain cannot comprehend. A single AI neural network can process and calculate at speeds that combined capabilities of humans can’t. In such cases, value of AI is derived from enabling entirely new capabilities or preventing high risks.
2.Possible tasks that are time consuming
Apart from entirely new range of capabilities, AI can undertake tasks that are expensive or slow if a human alternative is deployed. AI is merely a speed and accuracy booster for many jobs. The simplest example is an AI bot that can think just like a human chess grandmaster. Merely because of higher speed and depth of calculation, the AI can beat the greatest chess players of all time even though both follow a similar pattern recognition model to make their next move.
3.Repetitive tasks
AI is suitable for simple automated work where repeatable problems are solved at low marginal cost. AI is affordable, scalable, and delivers the most predictable ROI. In other words, AI can role-play as a complex network as well as a basic processing machine with equal ease.
Conclusion
Whether its return on AI investments by investors betting on AI wave or it’s enhancement of technical capability, the age of AI will be as disruptive as the dot com era. Many AI predictions will prove prescient and synthetic intelligence will replace human intelligence in most endevours. Whether investors earn an expected ROI in their portfolio or not depends on their entry and exit strategy and timing, the technical capacity of AI is definitely an ROI-positive undertaking.
