Advanced AI development is creating a financial loop where continuous investor capital masks the high costs of running large models [1].
This trend matters because the perceived affordability of AI may be an illusion. If the cost of compute, talent, and data remains prohibitively high, traditional businesses in the real economy may find the technology impossible to adopt profitably [2].
Industry analysts describe this phenomenon as a "circular cash loop." In this cycle, investors inject massive amounts of capital into developers like OpenAI and Anthropic, which then spend those funds on the immense compute power required to train models [1]. This process often obscures ongoing losses and the true operational expenses of the technology [2].
Some developers have already faced the financial toll of failed projects. OpenAI spent several months and millions of dollars [3] on a specific project before abandoning it. Meanwhile, other firms have withheld powerful models from the public. Anthropic recently withheld a model citing cybersecurity grounds, though critics said the move was also a tactic to manage investment hype [4].
Despite these systemic costs, some startups are betting that AI can be scaled for traditional industry. The startup Ciridae recently raised $20 million [5] in seed funding led by Accel. The company intends to build operating systems designed to bring AI transformation to real-economy businesses [5].
This creates a tension in the market. Some reports suggest the AI sector is currently in a bubble where revenue is far below the capital being invested [2]. Other perspectives suggest that companies should replace staff with AI now before the costs of the technology become even more prohibitive [6].
For now, the gap between the capital poured into AI labs and the actual revenue generated by real-world applications remains a central point of concern for economists [2].
“The high cost of developing and running advanced AI models is creating a financial “circular cash loop.””
The 'circular cash loop' suggests that the current AI boom is sustained by venture capital rather than organic profitability. If the cost of compute and data does not drop significantly, the technology may remain a luxury for the wealthiest tech firms, leaving small and medium-sized businesses unable to integrate AI without facing unsustainable operational costs.
