What if the artificial intelligence sector is a bubble?
The fear is that it could end up like the metaverse or the NFTs
Lifestyle
April 27th, 2024
April 27th, 2024
«Even if you believe AI will be today’s equivalent of electricity or the internet, we are at the very early stages of a highly complex multi-decade transformation that is by no means a done deal». This is stated by the Financial Times commenting on the transition period the artificial intelligence sector is going through. It cannot be called a "crisis," but experts argue that the field is scaling back after the initial enthusiasm. After about two years of growth, there is concern that the sector may turn out to be a major bubble. As reported by the Wall Street Journal, some companies that had bet on artificial intelligence are struggling to justify the huge initial investments. Recently, for example, the CEO of one of the sector's most popular startups, Inflection AI, resigned. In an interview with Bloomberg, he admitted that the company failed to find a sustainable business model. Developing AI-based platforms is very expensive, and the associated services still generate little profit proportionally. This is also why the sector is mainly dominated by companies with huge capital, such as Microsoft, Google, Meta, and Amazon, which, however, favors the concentration of power in a few entities, to the detriment of emerging ones.
Why AI is not yet economically sustainable
Despite their dominance in the sector, even the largest companies struggle to monetize with AI. The same goes for the most popular and user-appreciated companies, like OpenAI – which spends about $700,000 a day just to keep ChatGPT running, requiring immense computational power. Using AI platforms to perform very simple tasks (like summarizing an email) is like using a Lamborghini to deliver a pizza: according to Reuters, every request made to ChatGPT would cost the company four cents. Even GitHub Copilot, a virtual assistant capable of writing and checking computer code, despite its great success in the web development sector, is still economically unprofitable. The inability to reduce the management costs of such powerful AI services is pushing big companies to rethink their business models. Even Google is reportedly considering making a future AI-enhanced search service paid – a move that, if confirmed, would go against the company's traditional free approach.
The AI sector may undergo downsizing
After months of enormous attention and blanket investments, there is a growing belief in the sector that it is time to proceed with greater caution. A signal in this direction comes from OpenAI itself: in January, it introduced GPT Store, an online portal to download specific and customized versions of ChatGPT; the service was warmly welcomed by industry professionals, being likened to the equivalent of the App Store, which when introduced quickly revolutionized the market. A few months later, however, GPT Store failed to meet the high expectations, also because lately the company has been involved in numerous projects, such as Sora and AI-driven browsers, which have taken up space and attention from the service. The widespread suspicion among some analysts is that these are remarkable and potentially very useful tools, but that companies in the sector are overstating their actual application margins to avoid a decrease in investments. Will the excitement about artificial intelligence translate into a truly economically sustainable adoption? Or will the AI sector undergo downsizing like what happened with the metaverse and NFTs?
