When will half of highway miles be driven by self-driving cars? Will it be technologically possible for a robot to win a tennis match against a ranked human player, and if so, when? When will artificial intelligence have a measurable, positive impact on the median income of workers? When will AI significantly alter management practices within Fortune 500 companies?
AI and machine learning are already changing the landscape of labor and business, but many questions are still looming about the future. We used our March 8 MIT AI and Machine Learning Conference to gain further insight into a timeline for further changes by asking attendees for their views. We asked thought leaders, CEOs, and researchers when they expect major AI and machine learning milestones to be achieved.
The 16-question online survey asked participants about their beliefs on the future of automation, AI capabilities, and their effects on society. With 140 respondents we got a glimpse into the plausible – and the far-reaching – impact AI might have.
After a series of engaging keynotes and panel discussions – on business, skills and labor, innovation and entrepreneurship, as well as policy and governance – IDE Co-Director, Andrew McAfee, presented the survey results. Here are some of the highlights:
- The most consensus among participants regarded real-time translations and human-machine conversations (i.e. customer service transactions) as well as AI image interpretation (i.e. medical analysis) being routine within the next 8-10 years.
- The greatest disagreement was about events expected to happen further into the future, or never. Machines performing the majority of medical surgeries, for example, or being able to reproduce themselves, will not happen until 2034, if at all, according to respondents. The same held for factories having fewer than twenty workers.
- While participants doubted that improvements to AI will raise the median income of workers by 15%, a reduction in wage disparity seems most likely within the next 25 years.
Thanks to all the participants, and to Sebastian Steffen at MIT for preparing and analyzing the survey data.