What 81,000 people want from AI
Three factors that influence the adoption of AI.
Anthropic analyzed 80,508 open-ended interviews with Claude users from 159 countries and 70 languages. According to Anthropic, this is the largest and most multilingual qualitative study of its kind to date. The key finding may sound unspectacular at first: people primarily want AI for three things. They want to save time, do better work, and learn faster.
Precisely because this answer seems so obvious, it is strategically valuable.
After all, many AI initiatives don’t fail because of the technology itself. They fail because companies treat these motivations too abstractly. People talk about models, tools, and governance—but not enough about exactly how AI is supposed to be tangibly useful to people in their daily lives.
That is exactly where adoption begins.
The real insight is not futuristic, but human
People looking to implement AI are often searching for that big "aha" moment: the revolutionary feature, the tool with the biggest edge, the one use case that changes everything.
The Anthropic analysis shows something different. People aren't looking for an AI spectacle. They're looking for relief, quality, and guidance.
It’s not the novelty that attracts people. It’s the value.
Anthropic itself describes this in very concrete terms in many of its posts. A software engineer from Japan puts it this way:
"For the first time, I felt that AI had surpassed human capabilities in a business task. That day, I left work on time and picked up my daughter from daycare."— Software Engineer, Japan
This is a powerful example because it illustrates two things at once:
- AI improves the quality of work.
- The real value lies not only in the task itself, but in its impact on daily life.
This is precisely what is often underestimated in transformation projects. Companies tend to measure the success of an implementation based on adoption, usage, or tool availability. People, however, evaluate AI much more directly: Does it help me with my actual work? Does it improve my results? Does it free up my time or mental capacity?
Saving time is important—but only if it feels genuine
Saving time is one of the strongest reasons to start using AI. That makes sense. Anyone who can conduct research faster, prepare for meetings more effectively, condense documents, or speed up the creation of initial drafts will immediately see the benefits.
But saving time alone isn't enough.
Because if AI saves five minutes up front but creates twenty minutes of oversight down the line, it doesn’t build trust. Then productivity turns into friction. The Anthropic study also illustrates this backlash very clearly. A participant from Brazil describes it this way:
"I had to take photos to prove to the AI that it was wrong—it felt like talking to someone who just wouldn't admit they were wrong."— Employee, Brazil
This is an extremely valuable insight for leaders. Not because it is surprising, but because it highlights the real logic behind adoption: People don’t adopt AI because it works quickly. They adopt AI because it works reliably.
The key question, therefore, is not just:Where can AI save time?
Rather:Where does it save time without creating new uncertainty at the same time?
For many people, better work is the real key
A second key theme in the study is improving quality. People don't just want to work faster; they want to do better work.
This is particularly relevant for knowledge work. In many roles, the focus isn't on sheer volume of output, but rather on:
- more thorough preparation,
- clearer reasoning,
- more robust decisions,
- more structured communication.
The Anthropic analysis illustrates this through several high-level quotes. For example, a freelancer from the U.S. describes:
"The AI pieced together the historical clues and led to my correct diagnosis—after more than nine years of misdiagnoses."—Freelancer, USA
The example is extreme, but that is precisely why it is so revealing. The added value lies not in producing text more quickly, but in the ability to synthesize a large amount of information in a way that leads to a higher-quality assessment.
For businesses, this means that the most relevant AI applications are often not the ones that make the most noise. They are the ones that stabilize or improve quality.
A good example from everyday management would be:
- quickly turn lengthy project updates and meeting notes into a solid basis for decision-making,
- summarize complex customer or market information,
- Highlight counterarguments early on,
- or organize knowledge in a way that allows teams to act with greater confidence.
That’s exactly what we see in practice at Leaders of AI. The most powerful AI use cases are rarely the most spectacular ones. They are the ones that help people perform their actual work more accurately, quickly, and safely.
Learning isn't a side effect. It's a core benefit.
The third major pull factor identified in the study is particularly interesting: learning.
Many companies still view AI primarily as a productivity tool. However, the Anthropic findings show that people also use AI because they want to get up to speed on new topics more quickly, better understand how things are connected, and feel more competent.
An entrepreneur from Germany puts it very bluntly:
"I've probably learned more in six months than I did during my entire college career."— Entrepreneur, Germany
Of course, that’s a subjective statement. But that’s precisely where its value lies. It shows just how much of a game-changer AI can be when it comes to accelerating learning.
This is highly relevant for professionals and managers. Anyone working with AI today must do more than just speed up individual tasks. They must be able to familiarize themselves with new topics more quickly, better understand emerging trends, and actively shape their own role amid change.
Learning thus does not become a byproduct of AI—but rather one of its most important drivers of adoption.
What leaders and change managers should take away from this
The real management lesson to be drawn from the 81,000 interviews is therefore not that people like AI.
The real lesson is this: People embrace AI when it tangibly improves their work and development.
That sounds simple. Putting it into practice in the workplace is another story.
Anyone who wants to integrate AI must be able to provide specific answers to the following questions:
- Where does AI actually save us time in our daily lives?
- Where does it visibly improve quality?
- How does it help people become safer more quickly?
- Where, instead, does it lead to mistrust, the need for monitoring, or concerns about roles?
That is exactly what adoption work is all about.
Not primarily based on tool selection. Not primarily based on hype. But rather on how it translates into processes, communication, expectations, and empowerment.
Here’s an example: If a company announces AI as a general “co-pilot” but no one can explain exactly which specific work processes will benefit from it, the implementation remains abstract. If, on the other hand, it becomes clear that AI actually reduces the workload in areas such as analysis, preparation, knowledge work, or documentation, it becomes relevant.
The flip side is just as important: if unreliability, job-related worries, or a vague sense of losing control aren't actively addressed, stable usage won't take hold, even if there is interest.
The most important question regarding AI is not a technological one
The Anthropic study therefore serves primarily to correct a question that is often asked incorrectly.
Not: Which AI is currently the most powerful?
Rather: What kind of benefit becomes so valuable to people that it endures?
The real challenge for leaders isn't to roll out new tools as quickly as possible. The real challenge is to adapt these three drivers—saving time, doing better work, and learning faster—to their own organizational context.
This isn't just a peripheral communication task. It's leadership work at the heart of the transformation.
Conclusion: Adoption is not a technical issue, but rather a matter of translation
81,000 people, 159 countries, 70 languages—and in the end, it all comes down to three very human needs:
Save time. Do better work. Learn faster.
That is precisely where the real power of AI lies.
Not in the hype. Not in the tool comparison. Not in the next model generation.
Rather, the question is whether companies take these needs seriously enough to build a credible implementation around them.
People don't adopt technology just because it sounds impressive. They adopt habits that work.
Think ahead instead of scrolling on
Once you understand what people really want from AI, the next question is: How do you build a system that delivers exactly that—reliably, in everyday life, and without any friction?
That’s exactly what we teach in Master Business with AI MBAI): how to not only use AI, but also integrate it into your work in a way that ensures time savings, quality, and learning aren’t just empty promises—but become part of your daily routine. With a team of AI assistants that you build yourself. Practical, university-certified, and immediately applicable.
Sources
- What 81,000 People Want from AI — Anthropic, 2025
