AI Isn’t Replacing Work, It’s Refining Ego

What the Anthropic AI survey really tells us about ambition, identity, and modern work

Anthropic recently published one of the most revealing datasets yet on how people actually use AI in real life. They surveyed around 81,000 people using an AI interviewer to understand a simple but surprisingly under-explored question:

What do people actually want from AI?

Not in theory, but in practice. And the answer, perhaps unsurprisingly, is less about replacement and more about refinement.

People don’t want less work. They want to feel better at it.

The most common aspiration wasn’t to gain time back or escape from work, but instead professional excellence. In simple terms: people want AI to help them do their current jobs better. That alone reframes the entire narrative around artificial intelligence in the workplace, we often assume the end goal of AI is liberation from labour. But, the data suggests something much more interesting, that people are not trying to exit work, but upgrade their identity within it. In second place came personal transformation, followed by life management and financial independence.

The matter of AI being used as a ‘status amplifier’ and not a shortcut is particularly significant because it challenges a widely held assumption in AI discourse: that the primary value of AI is efficiency. In reality, efficiency is just the surface layer. What people are actually using AI for is capability elevation - writing better, thinking faster, producing at a higher standard. In other words, AI is becoming a tool for status within work, not just speed of work.

So, what is the wider significance of these results? Perhaps something uncomfortable about modern work culture, and the fact that people don’t want to finish early, but aim for mastery over margin. We want to be harder to replace.

The hidden outcome: raised expectations everywhere

Everytime AI improves output quality, the baseline shifts, what was once “good” becomes average, what was average becomes inadequate, so even as AI improves performance, it quietly raises the standard of performance required to feel successful. Personally, I feel that this result is not a relief, and it means that its an ongoing challenge to find your footing in the AI landscape. So what then should everyone focus on and how should we try to learn?

For anyone building with AI, whether from a creative industry or not the lesson is clear: the value of AI is not just in doing things faster, but in redefining what “good” looks like, and we are the ones with the power to decide that and direct it.

Which means the real question is no longer: How do I save time using AI?

But instead: What version of me is AI enabling me to become?