Professor Pedro Domingos said, "By 2023, artificial intelligence has added $5 trillion to the market value of the top 7 technology companies." These include Meta, Amazon, Apple, Alphabet, Microsoft, NVIDIA, and Tesla, the "Big Seven," which are driving stock markets worldwide in the name of artificial intelligence.

Not only large tech companies, but also the valuations of AI research startups are high, reaching millions (or even billions) of dollars. However, analysts have questioned their revenue models and doubt their long-term profitability.

"This is a classic example of a big market illusion," said Rob Arnott, founder of venture capital firm and Research Affiliates, who recently equated the hype around artificial intelligence to the dot-com era of 2000 when everyone was betting big on the internet, changing everything.

The reason AI companies receive high valuations is often due to their growth potential rather than current earnings. The return model of AI companies is based on future promises rather than actual sales. Interestingly, according to a study by Microsoft conducted through IDC, for every $1 invested in AI, the company can expect an average return of $3.5, and over 5% of organizations globally can even achieve an average return of $8.

VCs attribute the high valuations of AI to high usage rates of the products. They believe that the value can be demonstrated by converting free users into paying customers. However, this is not entirely accurate.

Large tech companies providing cloud services have injected millions of dollars into AI research companies, such as Microsoft's $13 billion investment in OpenAI and Amazon's $4 billion investment in Anthropic and Perplexity AI. The full details of their strategic partnerships in terms of revenue expectations are not yet clear.

Recently, valuations of AI startups have been skyrocketing. Mistral AI, a French AI startup founded just eight months ago, has surpassed a valuation of $2 billion.

While OpenAI is negotiating to raise investments at a valuation of $100 billion, Anthropic, an AI startup, is negotiating to raise $750 million at a valuation of $18 billion. In a little over a year, the valuation of AI-driven search engine "Perplexity AI" has reached $520 million.

Chamath Palihapitiya, CEO of Social Capital VC, said, "The AI industry is currently in a strange state." He believes that despite significant investments in capital expenditure, credit, and tokens, customer revenue does not seem to have increased significantly.

Palihapitiya also believes that AI primarily focuses on improving efficiency, leading companies to adopt agents and automation to reduce costs, enabling AI to save operational expenses. Therefore, this is a different way of evaluating AI companies.

Considering the development of the AI market in the past year, the companies in this field are relatively new. Therefore, comparing their revenue or profit models with existing SaaS or regular startups may be incorrect. It is too early to consider these investments as overvalued.

According to a recent report, the profit margins of AI startups are said to be lower than those of existing enterprise software companies, raising concerns about the overvaluation of these AI companies.

Unlike the expectation for startups to achieve profitability within three to four years and for SaaS companies to achieve profitability within five to seven years, the expectations for AI research companies to truly achieve profitability cannot follow the same path.

At the 2024 Davos Forum, Microsoft CEO Satya Nadella discussed the broader applications of artificial intelligence and shared how AI has become a growth driver for many economies. Nadella said, "There is no economic growth in the world after adjusting for inflation. This is a very disappointing state. Developed countries may experience negative economic growth." He added that he is optimistic about AI as a general-purpose technology that can drive economic growth.