Falaah Falaah AI
Insights

What 80,000 People Want From AI — And What It Means for Your Business

Anthropic's landmark study of 80,508 people across 159 countries reveals a surprising truth: people don't want AI to work faster. They want to live better. Here's what that means for small businesses.

FT
Falaah Team
· · 17 min read
What 80,000 People Want From AI — And What It Means for Your Business

Anthropic — the AI company behind Claude — just published the largest multilingual qualitative AI study ever conducted. They interviewed 80,508 people across 159 countries in 70 languages and asked a deceptively simple question:

What do you actually want from AI?

The answer wasn’t “make me faster.” It wasn’t “generate more content.” It wasn’t even “automate my job.”

People don’t primarily want AI to work faster — they want to live better.

— The study’s central finding

The dominant theme across 80,000 responses: take the busywork off my plate so I can focus on what matters. That’s not an AI story. That’s a human story. And for anyone running a small business, nonprofit, or professional services firm, it’s deeply personal.


The Study at a Glance

Before we dig in, here’s what makes this research exceptional:

DetailValue
Participants80,508 Claude users
Countries159
Languages70
MethodConversational AI interviews (not surveys)
DurationOne week in December 2025
AI delivery rate81% said AI already took steps toward their vision
Avg. concerns per person2.3 — people hold hope AND worry simultaneously
Global sentiment67% net positive

This isn’t a poll with five answer choices. These were open-ended conversations where people described their actual experiences, frustrations, and hopes in their own words. The depth of the responses is what makes the findings so valuable.


What People Actually Want

The study categorized responses into nine distinct visions for AI. The top five paint a clear picture:

What People Want% of RespondentsIn Their Words
Professional Excellence18.8%Handle routine tasks so I can focus on strategic, meaningful work
Personal Transformation13.7%Help me grow, understand myself, achieve wellbeing
Life Management13.5%Reduce the mental burden of managing everything
Time Freedom11.1%Give me my time back — for family, hobbies, rest
Financial Independence9.7%Help me achieve economic security

The remaining categories — Societal Transformation (9.4%), Entrepreneurship (8.7%), Learning & Growth (8.4%), and Creative Expression (5.6%) — round out the picture.

Here’s what jumps out: roughly one-third of all respondents want relief from time, money, and mental bandwidth constraints. Not more features. Not more capabilities. Relief.

“I didn’t want AI to make me a better worker. I wanted it to give me back the hours I was losing to tasks that shouldn’t need a human.”

— Paraphrased from study respondents

What’s Missing From This List

Notice what nobody said:

  • “I want a chatbot”
  • “I want to generate images”
  • “I want a smarter search engine”

The asks are deeply operational and deeply human. People want to stop doing busywork, stop feeling overwhelmed, and start spending time on what actually matters to them — whether that’s growing a business, serving a community, or being present with family.


Where AI Has Already Delivered

The study didn’t just ask what people want. It asked what AI has actually done for them. The results are encouraging — and revealing:

Where AI Delivered%What It Looks Like
Productivity32%Dramatically accelerated work, automated repetitive tasks
Cognitive Partnership17.2%Served as thinking partner and creative collaborator
Learning9.9%Helped acquire new skills and understanding
Technical Accessibility8.7%Enabled building things that were previously impossible
Research Synthesis7.2%Processed large volumes of information efficiently
Emotional Support6.1%Provided guidance and judgment-free space

But there’s a notable gap: 18.9% said AI fell short of their expectations. The technology has delivered real value, but it hasn’t yet closed the distance between what people want and what they’re getting.

That gap is the opportunity.


Why This Matters More for Small Businesses

Here’s where the data gets really interesting for anyone running a small or mid-size organization.

Independent Workers Are 3x More Empowered

The study found a striking occupational pattern: independent workers — entrepreneurs, freelancers, small business owners — report 3x higher AI empowerment than employees at larger organizations (47% vs 14%).

Why? Because when you’re running a small business, every hour of busywork is an hour you’re not:

  • Growing the business
  • Serving clients
  • Building donor relationships
  • Being present with your family

AI doesn’t just save time for independent workers — it fundamentally changes what’s possible. A solo operator with the right AI tools can handle operational complexity that previously required a team.

The Equalizer Effect

The study also found that AI optimism peaks in emerging markets — South America, Africa, South and Central Asia — where people see AI not as a threat but as an opportunity ladder.

One respondent, an entrepreneur from Cameroon, put it perfectly:

“I’m in a tech-disadvantaged country… AI did it in 30 seconds. It’s an equalizer.”

This resonates deeply with us. The organizations most underserved by technology — small businesses running on spreadsheets, nonprofits juggling disconnected tools, professional services firms drowning in manual data entry — are exactly the ones that stand to gain the most from purpose-built AI.

It’s also why we architected Muin for global reach from day one — multi-language support (English, Spanish, Arabic with RTL), multi-currency (33 currencies), regional data residency routing so your data stays in your geography, and WhatsApp as a first-class communication channel alongside email and SMS. When an entrepreneur in Cameroon says AI is an equalizer, we take that seriously. The infrastructure to serve them shouldn’t be an afterthought — it should be in the foundation.

Entrepreneurship as Capital Bypass

The study found that 8.7% of respondents see AI as a force multiplier for building businesses — and this vision peaks in Africa, South/Central Asia, the Middle East, and Latin America. People in these regions frame AI as a way to bypass the capital and infrastructure barriers that have historically kept small organizations from competing with larger ones.

That’s not a technology story. That’s an economic mobility story. And it demands software that’s ready to meet them where they are — in their language, their currency, and their region.


The Trust Gap: People Want AI But Don’t Trust It Yet

The study identified 13 categories of concern. People aren’t naive about AI’s risks — on average, each respondent mentioned 2.3 distinct concerns. The top five:

Concern%The Worry
Unreliability26.7%Hallucinations, inaccuracy, burden of verifying AI output
Jobs & Economy22.3%Displacement, unemployment, growing inequality
Autonomy & Agency21.9%Loss of human control, forced adoption
Cognitive Atrophy16.3%Losing skills, intellectual passivity, critical thinking decline
Governance14.7%Lack of legal frameworks, insufficient oversight

Other significant concerns include misinformation (13.6%), surveillance and privacy (13.1%), malicious use (13%), and meaning/creativity loss (11.7%).

The #1 Concern Is Solvable

Unreliability — hallucinations and inaccuracy — tops the list at 26.7%. It’s also the only “tension” in the study where the negative experience outweighs the positive.

Here’s the thing: this concern is largely driven by general-purpose AI tools that are designed to do everything and therefore improvise on everything. When you ask a general chatbot to process an invoice, classify a compliance document, or route an approval workflow, you’re asking it to wing it.

Purpose-built AI — AI designed for specific business tasks within defined guardrails — is fundamentally more reliable than general-purpose chat. Structured extraction pipelines with validation, confidence scoring, and human-in-the-loop review for edge cases are different in kind from “ask the AI and hope for the best.”

The study validates what we’ve seen firsthand: the path to AI trust isn’t better chatbots. It’s AI embedded into business operations where it can be verified, validated, and held accountable.

Privacy Is a Top Concern

13.1% of respondents worry about surveillance and data exploitation. This concern is highest in North America (17%) and Oceania — the very markets where most AI tools are built.

For small businesses handling sensitive data — financial records, employee information, donor details, client contracts — this concern is not abstract. It’s operational.

The architecture of your AI matters. Where your data goes, who can access it, and whether it’s used to train models that benefit someone else — these aren’t philosophical questions. They’re business decisions.

At Muin, we chose AWS Bedrock specifically because of its contractual guarantees: your data is never used to train AI models, and it never leaves our secure infrastructure. That’s not a policy that could change — it’s architecture.


Five Tensions: What the Study Reveals About Living with AI

The most nuanced finding in the study is that benefits and risks aren’t experienced by different people — they’re experienced by the same people simultaneously. The researchers identified five key tensions:

1. Learning vs. Cognitive Atrophy

  • 33% mentioned AI helped them learn; 17% worried it would erode their skills
  • 91% of those who cited learning benefits had experienced them firsthand
  • Only 46% of those worried about atrophy had actually witnessed it

What this means for business: AI should automate operational tasks (data entry, routing, classification) — not judgment tasks (strategy, relationships, decisions). The goal is to free up human capacity, not replace human thinking.

2. Better Decision-Making vs. Unreliability

  • 22% valued AI for decision support; 37% worried about reliability
  • This is the only tension where the negative outweighs the positive
  • Lawyers report both the highest benefits AND the highest reliability concerns (nearly 50%)

What this means for business: AI-generated insights need to show their source data. Confidence scores, citation trails, and human approval workflows are essential — not optional.

3. Time-Saving vs. Illusory Productivity

  • 50% cited time savings; 18% worried about losing time to AI or accelerating pace demands
  • Self-employed workers are most likely to experience both sides simultaneously
  • Only 1% have actually experienced illusory productivity — 17% just expect it

What this means for business: Real AI productivity doesn’t just speed up tasks — it eliminates entire categories of work. You don’t process invoices faster; the invoices process themselves. That’s the difference between “faster treadmill” and “getting off the treadmill.”

4. Economic Empowerment vs. Displacement

  • 28% cited economic benefits; 18% feared displacement
  • Independent workers report economic benefit at 3x the rate of institutional employees
  • Freelance creatives experience the tension most acutely: 23% benefit, 17% displacement

What this means for business: AI platforms should empower small organizations to compete — giving a 5-person team the operational capacity of a 50-person organization. That’s empowerment, not displacement.

5. Emotional Support vs. Dependency

  • 16% found meaningful emotional support from AI; 12% feared dependency
  • People who value emotional support are 3x more likely to also fear dependency
  • A Ukrainian soldier noted that AI companions helped during wartime when human support was unavailable

What this means for business: AI in business should be a tool, not a companion. It manages your operations so you have more time for human relationships — not less.


The Global Picture: AI Optimism by Region

One of the study’s most fascinating findings is geographic. AI sentiment varies dramatically by region:

RegionNet SentimentKey Driver
South AmericaMost positiveAI as economic opportunity
Sub-Saharan AfricaMost positiveAI as equalizer; 18% expressed no concerns at all
South & Central AsiaHighly positiveLearning vision 13-14% (vs 8% globally)
East AsiaPositive with nuancePersonal transformation peaks at 19%
North AmericaMixedHighest governance (18%) and surveillance (17%) concerns
EuropeMost cautiousHighest overall concern rates

Why the divergence? In emerging economies, AI is framed as an opportunity ladder — a way to bypass the capital, infrastructure, and institutional barriers that have historically kept small organizations from accessing enterprise-grade capabilities. In developed economies, AI is more often framed as a threat to established systems and employment.

This maps directly to the SMB opportunity. For organizations that have been underserved by enterprise software — whether a growing business in Lagos, a nonprofit in Louisville, or a professional services firm in London — an AI-powered platform isn’t a disruption. It’s an invitation to finally compete on a level playing field.


What This Means for Business Software

The Anthropic study confirms something that many of us in the small business world have felt intuitively but couldn’t prove with data:

The AI products that will win are the ones that deliver operational relief — not novelty.

People don’t want another tool to learn. They want fewer tools that do more. They want their weekends back. They want to stop worrying about the grant report that’s due, the invoice that was processed wrong, the compliance certification that expired without anyone noticing.

That’s what we’re building at Falaah AI with Muin — and the study data maps directly to the capabilities we’ve built.

From Study Findings to Real Capabilities

Here’s how each finding connects to what Muin actually does:

“Handle routine tasks so I can focus on strategic work” (18.8%)

This is the job of Muin Agents — AI-powered workers that handle repetitive operations 24/7 without human intervention. The Invoice Agent reads, extracts, validates, and books invoices automatically. The Compliance Agent monitors regulatory changes and flags policy gaps before they become audit findings. The Certification Agent tracks expiration dates across your entire team and sends multi-tier alerts before deadlines hit.

These aren’t chatbots. They’re purpose-built AI workers that do the work — not just answer questions about it.

“Reduce the mental burden of managing everything” (13.5%)

The mental burden comes from juggling disconnected tools. QuickBooks for accounting. Gusto for payroll. Google Drive for documents. Spreadsheets for everything else. None of them talk to each other — and you’re the human integration layer holding it all together.

Muin connects and unifies that patchwork into a single platform: Document Intelligence for reading, extracting, and creating documents. Smart Forms for data capture with conditional logic and workflow triggers. Integration Hub to sync with the tools you already use (HubSpot, QuickBooks, Google Drive, Gusto, and more). One login, one source of truth, one place to look.

“Give me my time back” (11.1%)

This is the difference between faster and gone. Workflow Automation doesn’t just speed up your approval chains — it runs them automatically. A visual builder for multi-step processes: when an invoice arrives, it’s extracted, matched to a PO, routed for approval, and synced to your books. When a certification expires, the renewal workflow triggers. When a new hire signs their offer letter, onboarding kicks off across HR, payroll, documents, and training — without anyone staying late to set it up.

The study found that only 1% of people have experienced illusory productivity from AI — the worry that AI just makes you faster on a treadmill that keeps speeding up. That’s because the best AI doesn’t make you run faster. It takes you off the treadmill entirely.

“AI as a force multiplier for building my business” (8.7%)

The Muin Assistant lets you chat with your business data in natural language. “Which vendors have expiring insurance certificates AND outstanding invoices over $5,000?” — a question that’s impossible to answer when your data lives across six tools, but instant when it’s unified. Smart Payments lets you accept cards, NFC, ACH, and Apple Pay (via Stripe) — and send payments via Stripe Connect — with a transparent 0.5% platform fee. No stitching together three payment tools. No reconciliation nightmares.

For a solo operator or a small team, these capabilities mean competing with organizations ten times your size — without ten times the staff.

“AI as an equalizer” (peaks in emerging markets)

The Communications Hub unifies WhatsApp, email, and SMS into one inbox — with AI designed to draft and handle routine responses automatically. A congregant sends a WhatsApp message asking about donation receipts? AI responds with the right information, in the right language, instantly. A vendor emails asking about payment status? AI drafts the response from real data — you just review and send.

For organizations where staff time is the scarcest resource — a small nonprofit, a growing professional services firm, a house of worship with one administrator — AI-powered communications aren’t a luxury. They’re the difference between responsive and overwhelmed.

Addressing the Concerns

The study’s top concerns map to deliberate architectural decisions in Muin:

Concern (Study)How Muin Addresses It
Unreliability (26.7%)Purpose-built AI for specific tasks — structured extraction, confidence scoring, human approval workflows. Not a general chatbot improvising on your business data.
Privacy (13.1%)Built on AWS Bedrock — your data never trains AI models, never leaves our infrastructure. Full audit logs. 256-bit encryption. Privacy-first by architecture, not by policy.
Autonomy (21.9%)Human-in-the-loop on every decision that matters. AI handles the operational work; you keep control of strategy, relationships, and judgment calls.
Cognitive Atrophy (16.3%)Muin automates data entry, routing, and classification — not thinking. Your expertise stays sharp because you’re using it on the work that actually requires it.

This isn’t a generic AI tool hoping to be useful. It’s a connected platform with business modules for Finance, HR, Compliance, and Non-Profits — with more on the way — where every capability is designed to deliver the exact operational relief that 80,000 people just told Anthropic they’re looking for.


The Bottom Line

Anthropic’s study asked 80,000 people what they want from AI. The answer was remarkably human:

Help me focus on the work that matters. Give me my time back. Take the busywork off my plate.

Not “make me a better prompter.” Not “generate more stuff.” Just: let me do less of what drains me and more of what fulfills me.

For small business owners, nonprofit leaders, and professional services partners, that’s not an abstract wish. It’s the difference between leaving at 5:30 and leaving at 8:00. Between growing the practice and just surviving. Between serving the community and drowning in paperwork.

The technology to deliver on that wish exists today. The question is whether it gets built for the organizations that need it most — or only for the ones that can afford six-figure implementation budgets.

We know which side we’re building for.


Sources