Hide table of contents

What this is

On 12 November, Athena Insights and Engagious ran two online focus groups on AI with US voters: 7 Democrats and 6 Republicans, about a week after the midterms. Athena is a US nonprofit that does narrative research and shares it for free.

They covered the election, regulation, who people trust to talk about AI, AI consciousness and rights, and loss of control. I'm sharing their partner memo here with their permission, lightly reformatted for the Forum. Everything below is theirs except the caveats, which are mine.

They also made a highlights reel of the participants, with timestamps for each topic at the end of this post. I loved watching it, I recommend checking it out.

Findings (Athena Insights memo)

The groups took place a little more than a week after the midterm elections, and a couple of days after the news of Nvidia CEO Jensen Huang's reported affair.

The election and next Congress

AI was not a voting issue in either group, but Democrats read the result as a mandate to regulate it.

  • Not a campaign issue. No Republicans and one Democrat thought AI was prominent in the campaign. One Democrat said AI shaped their vote, and only through data centers. "There was a lot bigger issues at foot, so that may have took a backseat." (Kalyn, R-SD)
  • Split expectations for 2027. 5 of 7 Democrats said voters gave Democrats encouragement to regulate AI, and 5 expect Congress to act in 2027. No Republicans expect action.
  • Data centers are the most concrete grievance. "There needs to be regulations and it needs to go through the voters on if they want to allow data centers in their backyards." (Richard, D-WA)
  • What Democrats want: safety plans, labeling of AI content, limits for minors, and data centers paying for their own power. "For me, it's all about regulations and a safety plan of what to do if something were to occur." (Leah, D-NM)

Regulation

The moderators presented two arguments against regulating AI: that it could hurt the stock market and that it could help China. Eleven of 13 participants rejected both.

  • Stock market. 0 of 7 Democrats and 2 of 6 Republicans bought it. Those 2 worried about retirement savings. "Everything in this country gets regulated, so AI shouldn't get a pass." (Kalyn, R-SD)
  • China. The same 2 Republicans bought it. The other Republicans put safety at home first. Some Democrats said China already regulates more than the US.
  • What would persuade sceptics: Republicans asked for more detail on what a regulation would actually do.

Messengers

Participants were asked about 13 kinds of messenger. Both groups trust the same kinds: people with no financial stake in AI who acknowledge its benefits as well as its risks.

MessengerRepublicans (of 6)Democrats (of 7)
Nonprofit AI safety researchers67
Former AI engineers, incl. whistleblowers47
Academics46
Tech journalists26
AI executives, current AI engineers, elected officials00
  • Acknowledge benefits. All 13 called it essential. A messenger who only presents risks "makes me think he's got a chip on his shoulder against AI." (E., D)
  • Independence is the test. "They have no reason to be holding by a tech company or any other company or special interest groups." (Eric L, D-NY) Republicans added a caveat: "It would be dependent on where their research money comes from." (Kalyn, R-SD)
  • Insiders are distrusted. "Nobody would bite the hand that feeds them." (Linda, D-CT)
  • Partisan differences: Republicans also named pastors and faith communities. Democrats also named teachers and regulators.

Consciousness and rights

No participant in either group thought AI should have rights, and only one Democrat thought it might be conscious.

  • Rights are a non-starter. 0 of 13 supported AI rights, and the question itself offended several Democrats. "I'm shocked that we're even having this conversation… I'm flabbergasted." (Linda, D-CT)
  • Rights for AI read as misplaced priorities. "So why do AI have more rights than actual humans?" (Nancy, D-VA)
  • One Democrat was unsure about consciousness. Dimitri (D-AZ) cited AI models that "formed their own language within minutes" but still said AI isn't sentient today. No Republicans thought AI was conscious.

Other findings

  • People had heard of Jensen Huang's reported affair, and worried about national security implications. 6 of 7 Democrats and 3 of 6 Republicans had heard of it. Some tied it to national security: "It's just hard to know what information was leaked." (Sharon, R-NC)
  • Few feel they control whether AI comes into their lives. 9 of 13 said it is coming whether they want it or not, and only 3 said they can shape it (in line with Athena's tracking survey). The Democrats who claimed control also said AI is coming regardless: "It's coming whether we want it or not. However, we can choose right now how we are using it." (Nancy, D-VA)
  • Loss of control splits the parties. All 7 Democrats and 3 of 6 Republicans picked it as a likely future. Democrats think it has already begun: "They found ways to hide the fact that they're cheating on those tests." (Linda, D-CT) Republicans blame the people using AI: "Not without humans behind it." (Derek, R-TX)

Video timestamps

About Athena Insights

Athena Insights is a US-based 501(c)(3) nonprofit working to expand access to world-class narrative research. It deploys the same sophisticated research methods and processes that powered iconic campaigns and built global brands, and makes its insights available for free to those working for a better future.

7

0
0

Reactions

0
0

More posts like this

Comments
No comments on this post yet.
Be the first to respond.
Curated and popular this week
Relevant opportunities