In March 2026 (First quarter), an AI-generated article appeared on Fortune exclaiming that the sky is indeed falling. The Littles have been predicting this since artificial intelligence (AI) large language models (LLM) burst on the scene in 2022.
The world was unprepared then, and many of us remain unprepared now. I know people who refuse to use LLM. I myself refuse to risk their use in any production process. Others have already essentially yielded to their new overlords. I know people who cannot decide if they want fries with that without asking ChatGPT.
The recent article was written by journalists using an LLM and verified by an editor. That is clarified in a note at the end of the story. The phrasing of "used" an LLM is interesting, and makes no real commitment regarding how the article was written or by whom. That is for another day.
The revelation of this piece is that four years on, there is "an AI breakthrough ... coming in 2026." Fortune, the author, or the LLM (hereafter "they") contend that this will be "a massive AI breakthrough," and it will arrive before July 1, 2026 ("first half of 2026").
The prediction is for a "transformative leap in artificial intelligence," based upon the input of "an unprecedented accumulation of compute." The fundamental is that more and more analysis, training, and processing is being directed at the AI tools and their evolution.
Elon Musk is quoted as believing that this volume of investment "will effectively double a model’s 'intelligence.'" This is expected to drive continued demand for electricity and related resources such as natural gas for turbines, and reallocation of resources from other computing functions (bitcoin mining is a noted example).
The end result is anticipated to be increasingly functional "thinking" machines. They expect these devices to increasingly be capable of multi-step processing and functions, "at or above the level of human experts on economically valuable tasks."
This has significant implications for the integration of these "thinking" machines into workflow and workplaces. They believe that there will be "large-scale workforce reductions because of AI efficiencies," which they believe have already begun in earnest.
The end result will be "a powerful deflationary force." That is an economic pressure that results in a general and sustained decrease in prices. The decreased prices could result in higher demand, or the potential exists for a more systemic economic downturn.
The downturn comes if the demand for AI-produced services or products remains relatively constant, but the related economic activity (spending) is diminished by the new influence, in this case, technology.
There is also the potential effect where economic displacement, like unemployment, either diminishes consumption or prevents any increased demand that might otherwise be encouraged by decreasing prices.
The Fortune predictions are staggering in terms of productivity leverage from LLM. One quoted CEO posits the possibility that significant enterprises may enter the competitive marketplace with a mere handful of human workers using vastly leveraged technology tools.
And, to think, much of this is predicted to happen in the next month or so. To whom will the displaced workers turn? If there are no jobs, will there be any consumption to drive the success of these AI-leveraged producers? If that scares us, we should focus on the article's predictions that even bigger changes in technology and capability may come as early as 2027.
They conclude that "The explosion is arriving faster than almost anyone is prepared for." If you are not seated with your seatbelt fastened, it may be time to do so. It may very soon become a bumpy ride for most of us.
