If you were born before 1975 or thereabouts, it should be a scene which you have witnessed before, just in a different garb of a new technological discovery cycle. This time, artificial intelligence is playing out in front of us. It was the Internet bust in the millennium.
Innovation by tech product creators powered by cheaply available venture capital has hyped up the capabilities of artificial intelligence to unrealistic highs and driven valuations, even higher. The usefulness, efficiency, practicalities, and risk impacts on our lives of artificial intelligence in everyday systems and their viability with respect to costs or profits have not figured at all in this entire dialogue.
Governments are fearful of stepping in to check the risks which are plainly in sight due to the technological race for global power and economic dominance. Since November 2022 when Chat GPT debuted, the fear of lagging behind has dominated rationality, pragmatism, and the need for peaceful coexistence. Even the fear of creating an unlimited technological destructive power (like Mythos) does not seem to have fazed the world.
In 2026, between $660 bn and $690 bn of total capital expenditure and AI infrastructure investments being made by Amazon, Alphabet, Meta, and Oracle1 has finally ensured that the day of reckoning for this hubris will come in the next six months. For the last three and a half years users of AI products did not have to pay for the older versions of AI products. There has been no connection or justification with actual costs and losses for the pricing of latest version of AI products by these product creator companies. The companies just picked a number out of thin air and announced it. Why would they want to do the arduous work of estimating real use levels and costs when it was so easy to talk up artificial intelligence capabilities, and drive-up valuations especially when cheap capital was flooding the market? Pricing did not seem to matter in such a scenario.
Even today, no one has tried to predict the actual scale of safe utility and adoption of AI in business and personal life which is a measure of how useful the technology will be. Without such studies it is not possible to calculate value, costs and profits and therefore informed price calculation becomes impossible. For the last three and a half years, winning more weekly users was all that mattered for these companies. Even the war in Ukraine and Iran was exploited by nations and AI companies like Palantir to drive up AI valuations. After all global power, technological superiority and economic strength was all that mattered. Right?
Wrong. The chickens are now coming home to roost. For some months now, companies have been demanding the adoption and use of AI by their employees. The amounts of AI token use mattered, the more the better. It became a core metric by which employee performance was measured. If you did not use more AI tokens at work, you were performing poorly. As a result, companies like Uber found that their annual token budgets were consumed in months (April 2026 for Uber forcing a re-evaluation of internal AI use policies). Baseless pricing to entice users to adopt their products combined with staged valuation growth and ill understood business models contributed to this debacle. Companies which had run out of budgets found themselves unable to define justifiable token usage policies for their employees. The confusion remains.
The final card is now being drawn by the major players like Anthropic and Open AI which are going for IPOs. The is the time when VCs can make a clean exit. This milestone marks the return of business logic and common sense to the AI growth story after a runaway years of delirium.
Business model transparency, revenue growth, and profitability will be measured from here on by public markets. That is why prices have been raised by many technology companies including Anthropic and Open AI recently. Microsoft has also raised prices for GitHub Enterprise Copilot. This pushes customers into a bigger quandary, just when their budgets have run out token prices are going up.
This is where the market mechanism of finding price equilibrium comes into play. The demand for AI’s true usefulness will become apparent over the next year. Supply must match demand in a fair market. That is when we will see billions in capital vanish. Most assuredly, the people who buy into these IPOs will lose money. That is the brutal reality of the market.
We will also see that some employees laid off in many areas will return to work. AI is notorious for being over-hyped even in coding. When it comes to maintaining and integrating the code we will need human coders too. More code is being written, but a disproportionate amount of it is not sticking2. The understanding of business or a user’s unique requirement and perfection in delivering it are yet far from being automated even by AI. Till that happens, in business we will continue to be needed. If we do permit AI to wield powers which enable it to replace us in the future, we will deserve our collective fate. As of now, many retrenched employees will return to work albeit in modified emerging roles.
This is the moment for Governments to enter and regulate AI. Technology product creators must be held accountable for the human risks impacts of their products. AI risk, human safety and security must take centre stage. The guiding of technology innovations is emerging as a pivotal role for governments to ensure peace, harmony and safety in increasingly knowledgeable and empowered national societies.
References:
- https://futurumgroup.com/insights/ai-capex-2026-the-690b-infrastructure-sprint/
- https://techcrunch.com/2026/04/17/tokenmaxxing-is-making-developers-less-productive-than-they-think/
Disclaimer: The opinions expressed in this article are personal opinions and futuristic thoughts of the author. No comment or opinion expressed in this article is made with any intent to discredit, malign, cause damage, loss to or criticize or in any other way disadvantage any person, people, company, government, country or global and regional agencies.