WASHINGTON—People are carefully cut up on whether or not they suppose synthetic intelligence (AI) will make issues higher, worse, or don’t have any influence on their very own lives and careers. The nation can be evenly divided on whether or not it is going to be good or dangerous for society and the subsequent era—however only a few suppose it can don’t have any influence, in keeping with a brand new survey carried out by Public First in partnership with the Middle for Knowledge Innovation.
The nationally consultant survey of over 2,000 U.S. adults was carried out earlier this month to trace how public attitudes and perceptions of AI have shifted since Public First carried out the same survey final 12 months. The emotion People expressed most regularly about AI in each surveys is curiosity. However that has ticked downward since final 12 months, whereas emotions of fear, nervousness, and concern have elevated.
“In our polling, we noticed that People had observed the latest accelerating enhancements in AI, however they had been additionally deeply divided about whether or not we needs to be extra excited or fearful,” mentioned Vinous Ali, Managing Director at Public First. “Whereas some People had been eager to experiment with AI tutors, coaches, and even relationships, others pointed to the threats from misinformation, automation and superintelligence. Getting the coverage and regulatory setting is essential however so too is constructing understanding of the way to make finest use of what AI might supply.”
Almost 6 in 10 People (59 p.c) suppose it’s seemingly that AI will improve unemployment. But lower than one-third of American employees (28 p.c) say they’ve used an AI chatbot at work, and, amongst those that have, 68 p.c say it has been useful and 38 p.c say it has change into a necessary device they use commonly. In the meantime, 70 p.c of People would help utilizing AI to diagnose well being circumstances if it’s double checked by a human, and 43 p.c of American males underneath 35 even categorical curiosity in having an AI girlfriend or boyfriend.
Among the many largest considerations, 71 p.c of People concern AI-generated content material might manipulate elections (a priority shared virtually equally amongst Democrats and Republicans).
In relation to AI coverage, 60 p.c of People say authorities ought to transfer cautiously earlier than creating new legal guidelines and rules to keep away from creating unintended penalties. When requested to decide on between two coverage priorities—both staying on the technological frontier or growing AI responsibly, even when it permits nations like China taking the lead—opinions are divided, with 43 p.c of People preferring to remain on the frontier and 35 p.c preferring accountable improvement.
“Policymakers have their work minimize out for them, not solely to design the appropriate insurance policies for AI and preserve the USA on target within the world AI race, but in addition to construct public help for these initiatives,” mentioned Daniel Castro, director of the Middle for Knowledge Innovation. “This job is very difficult as a result of AI and the businesses that make it are sometimes vilified within the media. As extra folks use AI, expertise the advantages, and notice their worst fears haven’t come to cross, hopefully public help will observe.”
The Middle and Public First are releasing the brand new survey findings at a briefing occasion that can be livestreamed from Capitol Hill in Washington, DC, Tuesday at midday Japanese. Extra findings observe.
As we speak:
- On the query of whether or not AI will make issues higher or worse for People personally, 32 p.c of U.S. adults say it can make issues higher, 22 p.c say it can make issues worse, and 33 p.c say it can don’t have any influence.
- When the query turns as to if AI can be good or dangerous for society as an entire, 38 p.c of People say it can make issues higher, 35 p.c say it can make issues worse, and simply 9 p.c say it can don’t have any influence.
- A 39 p.c plurality of American say they’re interested in AI on this 12 months’s survey, down from 42 p.c final 12 months.
- Amongst unfavourable feelings, 37 p.c say they’re fearful about AI (up from 32 p.c in 2023); 28 p.c say they’re anxious (up from 23 p.c); and 23 p.c say they’re scared (up from 19 p.c).
Alternatives:
- Giant majorities of People categorical curiosity in utilizing AI for a variety of private functions, from studying new data (73 p.c) to serving to cut back pointless power utilization (66 p.c), to managing well being and health (60 p.c).
- Majorities of People underneath 35 categorical additionally curiosity in an AI private assistant (62 p.c), tutor (51 p.c), exercise coach (51 p.c), or monetary advisor (51 p.c).
Considerations:
- 62 p.c of People are fearful concerning the potential influence of misinformation on the U.S. election, and 46 p.c lack confidence of their skill to detect faux AI-generated content material.
- 59 p.c consider AI will seemingly improve unemployment, and simply 7 p.c suppose it’s prone to improve employees’ wages.
- These with a bachelor’s or grasp’s diploma are barely extra prone to consider AI might change their jobs inside 20 years.
Coverage:
- Majorities of People help a variety of AI-related insurance policies on home points, from clear labeling of AI-generated content material (66 p.c) to banning using AI-generated content material for political campaigns (57 p.c), to holding AI corporations legally liable (64 p.c).
- In contrast, insurance policies which extra People noticed as a foul thought than a good suggestion had been introducing new taxes on AI fashions (32 p.c opposed, 24 p.c supported), banning new analysis into AI (35 p.c opposed, 22 p.c supported) and growing authorities funding for AI analysis (34 p.c opposed, 22 p.c supported).
“The problem that policymakers now face is completely encapsulated in the truth that public at the moment opposes extra authorities funding for AI analysis, which is on the prime of the coverage agenda that the Bipartisan Senate AI Working Group has simply launched,” mentioned Castro. “To stay a worldwide chief in AI, the USA wants to remain on the frontier of AI improvement, which requires sustaining a strong ecosystem of AI abilities, chips, knowledge facilities, and fashions. It additionally wants to guide in AI adoption, particularly in traded sectors of the economic system the place it faces world competitors, and areas like schooling and authorities the place there are important rewards. Each efforts would require intensive coordination and cooperation between the private and non-private sectors, in addition to a regulatory setting that fosters accountable innovation.”
Contact: Nicole Hinojosa, [email protected].
# # #
The Middle for Knowledge Innovation is the main world suppose tank finding out the intersection of knowledge, know-how, and public coverage. With employees in Washington, D.C. and Europe, the Middle formulates and promotes pragmatic public insurance policies designed to maximise the advantages of data-driven innovation in the private and non-private sectors. It educates policymakers and the general public concerning the alternatives and challenges related to knowledge, in addition to know-how traits corresponding to open knowledge, synthetic intelligence, and the Web of Issues. The Middle is part of the nonprofit, nonpartisan Info Expertise and Innovation Basis. For extra concerning the Middle, go to datainnovation.org.
Public First is an impartial, world coverage, technique and opinion analysis consultancy. It combines public opinion experience with deep data of the coverage panorama to ship data-rich, coherent coverage arguments which might be able to implement. Public First is an organization accomplice of the Market Analysis Society and members of the British Polling Council. For extra about Public First, go to: publicfirst.co.uk.
