With AI software increasingly holding the enterprise spotlight, businesses and investors are spending more than ever. In the first half of 2025, AI startups totaled over $44 billion for 2024. By the end of this year, Goldman Sachs analysis estimates that total investment in AI will rise roughly $200 billion.
But all that money is a reckless gamble to put it calmly. In the US at least, investors have essentially been betting farms on the idea that AI will quickly lead to increased labor productivity. This is a product of labor that can produce at a certain time, something that has never been seen in human history.
Despite the hype and intense, it's not happening. New reports by MIT researchers, first covered luckso far, we have found that 95% of attempts to incorporate generative AI into our business have failed.
According to a report entitled “genai divide: of ai of ai in business 2025,” MIT researchers found that only about 5% of companies that have succeeded in “rapidly accelerated revenues” were flattened.
Looking beyond the hype of AI marketing, it's not difficult to see why this is. Previous research on this type of AI has been sold as a kind of autonomous assistant, a white-collar worker, and found that the technology is not at the mark.
As of July, the best AI products are only completed 30% of the real-world office tasks assigned to them, but most have been significantly worsened.
Given that AI was previously expected to contribute to a global economy of over $6 trillion by 2030, the productivity gains from here must be nearly exponential to meet expectations.
For example, recent analysis MoneyWeek There's so much money on AI, and as we know, what runs through a complete upheaval of the world, claimed it knew it would look like a failure. A typical financial analysis shows that the top seven major tech companies should be seeing an extra $600 billion in annual revenue. (As of this year, they are projected to NAB about $35 billion a year, due to perspective.)
Every year, AI cannot return these constantly high numbers. The need for increased labor productivity increases – at least if Wall Street wants to justify the existence of spending, it can effectively kick half a trillion dollars.
Without a big breakthrough in the near future, it looks like a “when” issue, not a time when AI bubbles pop.
AI details: Companies that have replaced humans with AI are aware of their mistakes
