The nationally representative public opinion survey of U.S. adults sought public views about 15 different dimensions of people’s personal lives and major societal systems. Similar questions were asked in the non-scientific global canvassing of experts. The experts were also asked about 10 other aspects of life in addition to that list.

Comparing both samples, there are several patterns in the data. First, some key issues stand out as being relatively worrisome to both groups. For instance, both groups cited people’s privacy as the topmost area in which AI’s level of impact by 2040 will be far more or somewhat more negative than positive (see graphic below).

  1. Privacy – 66%
  2. Employment opportunities – 55%
  3. Politics and elections – 51%
  4. Relationships with others – 46%
  5. Basic human rights – 41%
  6. Levels of civility in society – 40%
  7. Wealth inequalities in society – 37%
  8. Physical and mental health – 35%
  1. Privacy – 79%
  2. Wealth inequalities in society – 70%
  3. Politics and elections – 67%
  4. Warfare – 61%
  5. Basic human rights – 54%
  6. Levels of civility in society – 52%
  7. People’s relationships with others – 46%
  8. Employment opportunities – 43%

(Percentages indicate participants who say the impact will be far more/somewhat more negative than positive.)

In addition, the public and these experts are relatively aligned in their views that AI’s effect on politics and elections, basic human rights, and the level of civility in society will be notably more negative than positive.

Second, though, there are also areas in which people’s opinions showed more worry than this sample of experts did. Those included AI’s probable impact by 2040 on people’s employment opportunities and the performance of K-12 schools and colleges. Conversely, experts expressed more concern than the public on AI’s effect on wealth inequality and people’s overall relationships with others.

It was notable, too, that there were times when these experts were somewhat more upbeat than the public. For instance, experts were more likely to say that the impact of AI would be that people would have access to knowledge and information from accurate, trusted resources. Experts also expressed more confidence than the general public about the potential impact of AI on the way people will spend their leisure time.

Finally, here is a rundown of the expert responses to some of the key questions that were not asked in the public opinion survey:

Warfighting: By a 61%-14% margin, these expertswere more likely to say AI would have a negative rather than positive outcome.

Criminal justice systems: Some 40% of these experts feel the impact of AI by 2040 will be more negative than positive, compared with 25% who feel the opposite.

Urban and rural life: Expert were notably more likely to tilt positive in their answers on these questions. Some 48% said AI would impact city life more positively than negatively and 40% said the same thing about AI’s effect on rural areas. Some 10% and 13% held negative views, respectively.

These experts were also considerably more upbeat than not about the coming impact of AI on the way people shop for goods and services and the way AI will affect transportation systems’ ability to move people and goods safely and efficiently.