VSand summer, investments in artificial intelligence (AI) jumped 81% globally. Seven unicorns were valued at more than $1 billion (€930 million), including five in generative artificial intelligence (GAI). AI raked in $25 billion in first half of 2023, says Crunchbase, including 10 billion invested by Microsoft in OpenAI, the designer of ChatGPT. Start-ups, often created by veterans of Big Tech, have flourished, announcing very innovative new products in a very short period of time. Big Tech is catching up and has launched powerful free software products in a very short time, including large language models (LLMs), the engines of AGI. Nvidia announced its superchip for AGI, the GH 200 Grace Hopper, for 2024. More than half of the companies, according to a Fortune-Deloitte survey, evaluate and experiment with AGI. 79% believe it will increase their efficiency, half (52%) believe it will increase their growth opportunities. More than a third (37%) are currently implementing it more or less.
In contrast, just under 40% of S&P 500 companies mention AI when presenting their financial results, and only 16% in their regulatory financial reports. In other words, its impact remains limited for the majority of companies.
Because the summer of 2023 also saw the emergence of pessimistic and critical opinions on AI, speaking of “generative AI-nxiety” (Harvard Business Review). Doubts remain about Silicon Valley’s ability to produce “the next big thing”, the next disruptive innovation, due to recent prior disappointments with the metaverse and cryptocurrencies. The financial results for the second quarter also showed that AGI did not yet have a clear impact, except for Nvidia, which is benefiting from its great technological lead and the shortage of AI chips.
Dystopian discourses
However, according to The Information, OpenAI is reportedly on track to generate $1 billion in revenue in the coming year. OpenAI currently generates around $80 million per month, after a loss of $540 million in 2022, and this summer launched ChatGPT Enterprise, a professional version of ChatGPT.
But investors remain concerned. Firstly, there remains significant regulatory uncertainty regarding AI, in the United States and in Europe. Then, having quality data is crucial for training AGIs, but the new requirement from rights holders to be paid for their data taken free of charge from the Internet by the creators of LLMs destabilizes them.
You have 51.01% of this article left to read. The rest is reserved for subscribers.