Financial Times essay says AI misses key parts of human listening

In short: A Financial Times opinion essay argues that AI can handle clear language well, but it does not capture how humans listen in uncertain, unspoken moments.

What's going on

Haru Yamada, a sociolinguist who studies how people use language, says AI is often compared to human intelligence in terms of speed and accuracy. But she argues that a bigger blind spot is listening, meaning how people take in what others say and what they do not say.

The essay says AI systems work by finding patterns in data and producing an answer based on what is already clearly defined. That can look like understanding, especially when the output is fluent. But Yamada says human listening is different because it includes tone, pauses, hesitation, mixed signals, and meaning that is still forming, like following a conversation while the speaker is still figuring out what they mean.

Yamada also points to work culture as part of the problem. Many workplaces reward quick, clear responses, so people learn to listen only for what is actionable and easy to label. She argues this can narrow what gets heard and can lead to miscommunication.

She uses “active listening” as an example. It can turn into visible behaviors, like nodding and repeating back words, rather than truly staying open to new information. She also notes that Japanese has different written forms for “to listen,” including one closer to functional intake of information and another closer to deeper engagement with what is not yet fully said.

What to watch

As AI tools spread in meetings, customer service, and everyday messaging, Yamada’s point raises a practical question. Will people start treating conversation more like filling out a form, with neat inputs and outputs, or will they protect space for slower talk where meaning can change midstream.

Source: Financial Times