Mike Litman
What 81,000 People Want from AI
Anthropic Research, March 2026

What 81,000 People
Want from AI

The largest multilingual qualitative AI study ever conducted

80,508
Respondents
159
Countries
70
Languages
Methodology

How they did it

Anthropic interviewed 80,508 people across 159 countries using AI-led adaptive conversations in 70 languages. All responses de-identified before analysis.

67%
expressed net-positive AI views globally
81%
said AI is taking steps toward their stated vision
Aspirations

Nine visions for AI

1
Professional Excellence
18.8%
2
Personal Transformation
13.7%
3
Life Management
13.5%
4
Time Freedom
11.1%
5
Financial Independence
9.7%
6
Societal Transformation
9.4%
7
Entrepreneurship
8.7%
8
Learning & Growth
8.4%
9
Creative Expression
5.6%
Reality

What AI has actually delivered

32%
Productivity Gains
Accelerated work completion and task automation
17.2%
Cognitive Partnership
Collaborative thinking and brainstorming
9.9%
Learning Support
Skill acquisition and subject mastery
8.7%
Technical Accessibility
Enabling previously impossible projects
7.2%
Research Synthesis
Information processing and analysis at speed
6.1%
Emotional Support
Judgement-free companionship and guidance
Concerns

Top 13 concerns

1Unreliability / Hallucinations
26.7%
2Jobs & Economy
22.3%
3Autonomy & Agency
21.9%
4Cognitive Atrophy
16.3%
5Governance
14.7%
6Misinformation
13.6%
7Surveillance & Privacy
13.1%
8Malicious Use
13.0%
9Meaning & Creativity
11.7%
10Overrestriction
11.7%
11Wellbeing & Dependency
11.2%
12Sycophancy
10.8%
13Existential Risk
6.7%

People aren't divided into optimists and pessimists. The same individuals experience both hope and fear.

Key Finding

Five tensions: light and shade

33%
Learning
vs
17%
Cognitive Atrophy
22%
Better Decisions
vs
37%
Unreliability
16%
Emotional Support
vs
12%
Dependency
50%
Time Savings
vs
19%
Illusory Productivity
28%
Economic Power
vs
18%
Job Displacement
Geography

Regional patterns

Most Optimistic
Developing Regions
Sub-Saharan Africa, South Asia, Latin America
  • View AI as an opportunity ladder
  • Focus on entrepreneurship and education
  • Less worried about jobs (17-18%)
  • Highest optimism scores overall
Most Cautious
Wealthy Regions
North America, Western Europe, Oceania
  • More cautious overall stance
  • Higher governance concerns
  • Privacy is a top priority
  • Job worry peaks at 24%
Most Reflective
East Asia
Japan, South Korea, China
  • Uniquely personal concerns
  • Cognitive atrophy worry at 18%
  • Meaning loss concern at 13%
  • Focus on individual impact

Meaningful AI deployment means addressing what "AI going well" looks like, grounded in actual user aspirations, not abstract risk discussions.

Source: Anthropic, March 2026
THE PUSHBACK
Where this breaks down

Self-reported aspirations don't predict actual behaviour. 81,000 people saying they want 'professional excellence' from AI tells us what they think they should want, not what they'll actually do. The gap between stated and revealed preferences is where most AI predictions go wrong.

The most honest finding might be that hope and fear coexist. Everything else is aspiration.

CONNECT THE DOTS

Anthropic asked 81,000 people what they want from AI. Jasmine Bina asked what AI is doing to us. One is the survey. The other is the diagnosis. Read them together and you see the full emotional landscape of the AI transition.

?
So What

What this means for builders

The people using AI most are also the people most worried about it. The companies that win will be the ones that take both sides seriously. The research is clear: people don't want cheerleading or doom. They want AI that delivers on their actual aspirations while honestly addressing their concerns.

Build trust through transparency Design for the tension Meet real aspirations
In Summary

Five things to remember

1

People want AI to make them better at their jobs, not replace them. Professional excellence is the top aspiration.

2

Productivity is the killer app. 32% cite it as the primary benefit they already experience.

3

Hallucinations are concern number one. Reliability is the single biggest barrier to trust.

4

Hope and fear coexist in the same person. The optimist/pessimist binary is false.

5

Where opportunity is lowest, optimism is highest. Developing regions see AI as a ladder, not a threat.

Mike Litman

Thank you.

mikelitman.me · hello@mikelitman.me

Continue from slide ?