(As part of this series, join MIT Technology Revieweditor-in-chief, Mat Honan, and editor-in-chief, David Rotman, for an exclusive conversation with Financial Times columnist Richard Waters on how AI is reshaping the global economy. Live Tuesday, December 9 at 1 p.m. ET. This is an event reserved for subscribers and you can register here.)
Will Douglas Heaven writes:
Every time I’m asked what’s next, a Luke Haines song sticks in my head: “Please don’t ask me about the future / I’m not a fortune teller.” » But there you go. How will things be in 2030? My answer: same but different.
There are huge differences of opinion when it comes to predicting the near-term impacts of generative AI. In one camp we have the AI Futures Project, a small donation-funded research group led by former OpenAI researcher Daniel Kokotajlo. The nonprofit organization made a splash in April with AI 2027a speculative account of what the world will look like in two years.
The story follows the meteoric progress of an AI company called OpenBrain (all similarities are coincidental, etc.) to a choose-your-own-adventure style boom-or-bust ending. Kokotajlo and his co-authors make no secret that they expect that over the next decade the impact of AI will surpass that of the Industrial Revolution – a 150-year period of economic and social upheaval so great that we still live in the world it created.
At the other end of the scale we have the team Normal technology: Arvind Narayanan and Sayash Kapoor, two researchers from Princeton University and co-authors of the book AI snake oilwhich push back not only most of AI 2027’s predictions but, more importantly, its fundamental worldview. That’s not how technology works, they argue.
Cutting-edge advancements can happen en masse and quickly, but changes across the economy and society as a whole move at the speed of humans. Widespread adoption of new technologies can be slow; slower acceptance. AI will be no different.
What should we think of these extremes? ChatGPT was released three years ago last month, but it’s still unclear how effective the latest versions of the technology are in replacing lawyers, software developers, or (gulp) journalists. And new updates no longer bring the incremental changes in abilities that they once did.