This year, America’s total spending to build data centers—the hardware behind artificial intelligence—will exceed the country’s construction spending on all commercial offices.
The automobile’s history offers a better model for understanding the future of AI than those often served up by the commentariat—such as the 1990s telecom boom and bust or the 1870s railroad expansion.
Consider: Engine-propelled vehicles were around for many decades prior to the launch of the 1908 Ford Model T, made affordable by the many and varied ingenious, incremental gains in inputs and processes that Mr. Potter maps. AI, similarly, has been available for decades (for specialists and in supercomputers).
But now the AI ecosystem, thanks to a similar pattern of ingenious gains, is exhibiting precisely the same efficiency and scaling trajectory as the automobile more than a century ago. The November 2022 release of ChatGPT was AI’s Model T moment, with consequential gains in efficiency still to come.
Read more | WALL STREET JOURNAL