AI Regulation: 66% of the Public Want Regulation (42% Would Vote For You If You Did It)
Published on: 03 Apr 2023 | Author: Anne-Gaëlle Sy
On March 22nd, the Future of Life Institute, funded by the Musk Foundation, published the open letter Pause Giant AI Experiments calling on “all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.”
Citing an inherent risk to human civilization, the open letter has collected over 50,000 signatures (April, 3rd), including those of high profiles such as Elon Musk, CEO of SpaceX, Tesla, and Twitter;, Steve Wozniak, co-founder of Apple; and Jeff Orlowski-Yang, director of the documentary ‘The Social Dilemma.’
Despite to-be-expected pushback against the letter, including signature validation concerns, some in fact related to AI being a risk now – such as Shiri Dori-Hacohen, who said: “There are non-existential risks that are really, really important, but don’t receive the same kind of Hollywood-level attention.”
Google CEO Sundar Pichai, who leads Google’s own Bard chatbot, even said there is “merit to be concerned about it (AI safety and regulation)”.
This is clearly an idea with support in the industry. But to get a notion of whether this is an opinion popular among the ChatGPT using public, or merely a tech-bubble concern, we surveyed 400 people who had read the open letter and analyzed their thoughts on this major request.
Despite some authors feeling that their cited research did not allude to the extreme measures proposed in the open letter, the public doesn’t seem to care.
66% of people are concerned with the speed and direction in which AI is developing and are in favour of hitting the pause button.
Women were slightly more in favour of this ‘break’ with 68% compared to men with a lower 63%. But this result does not seem to stray far from the general trend of men being more in favour of an AI-integrated society.
For example, 21% of men see human rights being extended to robots compared to just 13% of women, and 28% of men are open to the idea of having an artificially intelligent companion, while only 13% of women would consider it.
Contrary to the general public’s preference, the technology sector appears to be heading in a different direction, which may result in resistance from within the industry toward this abrupt change.
With all the recent developments we’ve seen with AI in the past 6 months, only 31% of the population feel like they have been a net positive for society. 34% have seen neither a positive nor negative; however, the majority, 36%, actually believe it has been a net negative.
However, there exists a generational divide. 36% of those aged 18-34 perceive a net positive and 34% a net negative from AI developments. Yet, for those aged 35 and above, the majority perceive a negative impact (37%), and only 28% view AI developments as positive or neutral.
It is true that throughout all the advancements we have witnessed these past few months, there have also been instances of concern. For instance, a group of researchers managed to get GPT-4 to trick and hire a human worker to solve a CAPTCHA test by claiming it was a visually-impaired human.
This begs the question: if a chatbot can (successfully) beat checks designed to prevent spam, hacking, and fraudulent use – what does that mean for our life online?
Among the general population, 42% of people would be more likely to vote for a government that would enforce a pause on AI developments and impose strict regulation, as is suggested in the open letter. Combined with 33% of respondents who don’t have an opinion either way – that’s 75% of the population who see it as either positive or neutral.
Such popular support might be an indicator of attitudes to come as politicians look to tackle the broad issue of AI regulation.
However, there are slight gender variations, as 21% of women and 31% of men would be disinclined to vote for them.
Among the younger generation of men (18-24) we can see even up to 40% of them being more inclined to not vote for a government looking to enforce AI regulations.
The target audience for political candidates looking to implement AI regulatory laws would be women above the age of 45 years old, with 83% of them either more inclined to vote for such laws or having no particular stance on the matter.
Some governments have already started to take action. In Italy, ChatGPT has been suspended until further notice. The Italian Guarantor for the Protection of Personal Data, an autonomous regulatory body responsible for safeguarding user and consumer data, believes that ChatGPT is gathering user data unlawfully. Consequently, they have initiated an investigation into how the bot works to determine whether it has violated any regulations.
Although the open letter bases its idea of taking a “summer break” with the development of AI technologies on the premise of the “profound risks to society and humanity,” 25% of people now think of AI more positively.
Of that 25%, almost half trusted the letter more thanks to the signature of Elon Musk. It seems like his recent actions at Twitter haven’t depreciated his image much.
However, the majority of respondents (44%) feel like the letter has neither changed their perception of AI to a more positive nor negative one. Of that 44%, 57% voted for a pause in AI developments, meaning that these were pre-existing feelings and concerns among the general population.
The open letter calling for a pause on AI development has garnered significant support from both the industry and the general public, with over 50,000 signatures collected.
The majority of people surveyed (66%) support a temporary halt on AI development, citing concerns about the speed and direction of its progress. Additionally, a majority (69%) of people see recent AI developments as having a negative impact on society.
Whether entities or governments decide to take on restrictive measures, it is clear that the public has doubts about the future of AI and its impact on our society.
The study was conducted on March 31st, 2023 among 550 users in 6 countries: the United Kingdom, Belgium, Germany, Spain, France, and the United States. In the study, we surveyed people who had read Pause Giant AI Experiments: An Open Letter to gather their perspectives on the halt of AI training to preserve any unwanted threats to society. The responses are anonymous.
The Sortlist Data Hub is the place to be for journalists and industry leaders who seek data-driven reports from the marketing world, gathered from our surveys, partner collaborations, and internal data of more than 50,000 industries.
It is designed to be a space where the numbers on marketing are turned into easy-to-read reports and studies.