Latest

Artificial Intelligence

The great AI gap: $1.4 trillion invested and $613 billion in revenue

AI has accumulated $1.4 trillion in investment against $613 billion in revenue, according to a counter that measures the gap in real time

Software is no longer bought, it is rented forever
Software is no longer bought, it is rented forever

The artificial intelligence industry has invested a cumulative total of around $1.4 trillion, while the revenue generated by those bets stands at around $613 billion, according to the data shown by isaiprofitable.com.The gap, of more than one trillion dollars, holds that most of the companies competing in the AI race have not yet recouped what they have spent on it. For a few months now, this site with a direct, almost provocative name has been circulating on the internet: isaiprofitable.com, something like Is AI profitable yet?.

The AI gap

Its creator, who prefers to remain anonymous, has built a counter that runs in real time, adding second by second how much money the big technology companies have invested in artificial intelligence and how much they have recouped in revenue.

Sovereign AI in Latin America: 11 projects in eight countries

The answer the page gives, for almost every player in the sector, remains the same: not yet. The proposal quickly gained visibility on social media and has been picked up by technology-focused outlets, which highlighted both the figure it displays and the way it presents it. It is a scoreboard that advances while the user watches the screen, as if it were the stopwatch of a race whose finish line is not yet in sight.

Amazon, the biggest investor

Among the companies analyzed, Amazon appears as the one that has devoted the most capital to this bet, with close to $313 billion invested.

At the other end is NVIDIA, identified by the site as the only major player in the sector currently making clear profits, thanks to its business of selling chips and specialized hardware, the input every other company needs to buy to train and run its models.

The site's creator himself qualifies the scope of these figures, saying they are not a verdict on the overall financial health of these corporations. Companies like Amazon or Google remain hugely profitable as a whole. What the scoreboard isolates is, specifically, whether their investments in artificial intelligence, separately from the rest of their businesses, have already paid for themselves.

The mechanics

The mechanics behind the counter are simple in design, though ambitious in their intent to convey immediacy. For each company, an annual rate of spending on artificial intelligence is estimated and that figure is divided by the number of seconds in a year, 31,536,000, which produces a spend per second. That value is what feeds the counter the user sees advancing on the screen.

The sources feeding these estimates combine official documents, such as the reports companies file with the U.S. Securities and Exchange Commission (SEC), with information obtained from quarterly earnings calls, financial leaks and calculations produced by outlets and specialized firms such as Bloomberg, The Wall Street Journal, The Information and Epoch AI.

The project's author himself has been explicit about the limitations of his method. In his own words, he does not claim these are audit-quality figures, but he considers the overall picture they paint hard to refute. It is an important clarification, since the site does not aim to replace rigorous financial analysis but to illustrate visually and forcefully a trend that many analysts had already been pointing out with less striking nuances.

Hasty interpretations

The outlets that covered the phenomenon agreed that the figure can lend itself to hasty interpretations. Seeing a gigantic number of accumulated losses growing in real time may suggest that the entire artificial intelligence industry is doomed to failure, when the reality is more nuanced.

This is a sector still under construction, where the big companies are deliberately investing in infrastructure, research and talent before the business reaches maturity, something common in high-upfront-capital technologies such as data centers or telecommunications networks in their early decades.

The site's creator himself seems to share that reading, since he argues that artificial intelligence is advancing faster than almost anyone anticipated and ventures 2030 as a symbolic deadline for finding out whether the so-called AI bubble ends up bursting or consolidating. It is a timeframe that, he explains, the industry itself has implicitly set through its projections and investment plans.

Within that same timeframe, it is mentioned that Anthropic would be approaching, according to the estimates gathered by the site, its first quarter with positive results. A sign that, although the sector as a whole remains in the red, some pieces of the machinery are beginning to show signs of profitability.

Alfredo Collosa

Author

Alfredo Collosa

Alfredo Collosa is a tax specialist. Public accountant. Bachelor's degree in Business Administration. Official Master's in Public Finance and Tax Administration from UNED/IEF in Spain. AMLCA Anti-Money Laundering Certificate (Florida University, US) and member of MetaLaw

The latest
From Huang to Botín, from Zara to Byron: corporate responsibility in global capitalism

Keep reading