What 2024 Has in Store for AI -I
OpenAI's head of sales was advised that the business would be secretly releasing a "low-key research preview," which wouldn't effect sales, just before ChatGPT was made public in November 2022. After more than 180 million users, it's safe to say that predicting the AI landscape is challenging.
However, not everything about ChatGPT's success was predictable. What happened a year ago would have been very impossible to predict: a dramatic boardroom expulsion at the world's most prominent AI business; an intensifying AI race between corporations and between countries; a U.S. Senate forum on the theme of "Doomsday Scenarios"
Predicting the future of AI is tough for the faint of heart due to the rapid growth of the technology and the diverse and wild reactions to it. However, TIME chatted with five experts who fearlessly expressed their predictions for the upcoming year in AI, despite being intimidated by the assignment.
Power-Deficient Data Centers
The scarcity of semiconductor chips turned into the first tangible sign of the AI explosion in 2023. According to Dan Hendrycks, executive director of the nonprofit Center for AI Safety in San Francisco, electricity demands will overtake all other factors in 2024.
Approximately 1% of the world's electricity is used by data centers. In Ireland, data centers consume nearly one-fifth of all electricity, a country that big IT businesses choose due to its low tax rates. AI presently uses about 20% of the capacity of data centers worldwide. In 2024, this percentage is probably going to rise significantly as AI systems continue to be trained and operated on ever-higher computing power.
Hendrycks thinks that businesses will attempt, and in fact are already attempting, to strike agreements with governments in order to secure a power supply. "To be receiving that level of electricity, you require the government's assistance in some way. Since these states' energy demands are only going to increase, some of these AI companies—I won't name them—will try to negotiate arrangements with their leaders regarding energy.
Oil-rich Middle Eastern nations, with their propensity to “pour money in weirder investments,” are expected to become more crucial in the global fight for AI supremacy, according to Hendrycks, since oil is one of the easiest ways to power data centers. Regarding the AI scenario for 2019, Hendrycks states, "In the way that we think of the U.S. and China as being relevant, probably the third most relevant [region] would be the Middle East."