What great speakers say about AI-made talks

In October 2025, on NPR's TED Radio Hour, Chris Anderson, introduced by the network as head of TED, made a prediction about the future of the stage. In the age of AI, he said, "people are going to want real contact with flesh and blood human beings". If you have a talk on your calendar, the line raises a practical question: what do the speakers, coaches and organizations that make a living teaching public speaking say about AI?

We gathered here the verified statements of six references in the field. They all arrive at a very similar point, and each one comes with caveats you need to know.

Chris Anderson: authenticity and your own voice on stage

In the same NPR interview, recorded at a time of leadership change at TED, Anderson said that "communication between humans is a very, very deep part of us" and said he believes this will not go away. For him, valuing our own humanity will matter more in the age of AI. The advice he gave speakers fits in one line: "be authentic, bring freshness, explain something in a different way, in your own voice".

The idea has been with Anderson for a long time. In 2013, in the article "How to Give a Killer Presentation", published by Harvard Business Review, he was already recommending "don't read PowerPoint slides" and "play to your strengths and be authentic". A talk built from a generic text goes in the opposite direction of this advice, because it puts words in your mouth that you have to memorize and learn to say as if they were yours.

Nancy Duarte: the efficient and empty presentation

Nancy Duarte, from Duarte Inc., published in April 2025 the piece "How Generative AI is Changing Presentations". In it, she writes that heavy use of AI "can atrophy the parts of the brain where out of the box creativity happens". She also describes the risk of an "efficient and empty" presentation and sums up the problem in one sentence: "Speed replaces empathy, and your message won't resonate."

One caveat matters here. Duarte treats AI as a collaborator and accepts using the tool throughout the text. Her concern lies in the place empathy holds in preparation. For Duarte, the human part of the work is where "you imagine, you connect and you move the room".

Carmine Gallo and the experiment with Melanie Perkins's story

Carmine Gallo wrote a guest piece in February 2026 for the Barnes & Noble blog, as part of the launch of his book Viral Voices. In it he describes a simple test. Gallo gave the beginning of the story of Melanie Perkins, founder of Canva, to three AI models and asked each one to continue the narrative. All three returned similar, predictable versions. The real story includes about 100 rejections from investors.

Gallo draws a conclusion from this about the role of the speaker. "AI has no human emotion", he wrote, adding that it has no personal stake in the outcome of what it produces. The speaker, according to him, is "the storyteller with a unique life experience". Gallo also recognizes the tool's value: for him, AI is "a powerful and useful tool to complement your story". Notice the word complement. In our view, it leaves the story in the hands of whoever goes on stage.

The detail of the 100 rejections shows why this matters. It is the kind of information that only shows up when someone has lived, researched or heard the story up close, and it is what makes the audience look up from their phones.

Julian Treasure and Brené Brown: being yourself in front of the audience

Julian Treasure took part in September 2025 in TED's How to Be a Better Human podcast. The episode does not mention AI and brings lines about authenticity that speak directly to the subject of this series. Treasure said that "the A is for authenticity, which means being yourself". He followed up with two short sentences: "be yourself. You are enough." And he added that pretending to be someone else never really works, unless you are an actor.

A talk written entirely by a tool puts you in exactly that actor's position, delivering a text another voice composed.

Brené Brown addressed AI directly at SXSW 2025. According to Fast Company Brasil, which translated her remarks, she said: "AI cannot have courage. Nor live stories that truly resonate". In September of the same year, in a keynote at the Workday Rising event reported by Fortune, Brené said the future will belong to those who can "balance the paradox between humanity and technology". Her position seeks balance, and the line about courage shows where the part that only a person can offer lies.

The National Speakers Association and the industry's trust problem

The National Speakers Association, the association of professional speakers in the United States, published in September 2026 a piece by Jaime Nolan on credibility in the speaking market. One line sums up the tone: "The speaking industry has a trust problem". The text states that this problem began before AI and that the tool only made it "cheaper, faster and available to anyone with a laptop".

According to the article, event organizers receive pitches from speakers whose pages look identical to one another and indistinguishable from AI generated profiles. The conclusion is that "visual polish is no longer a reliable signal of anything".

Read it with the caveats close at hand. The article promotes the CSP certification offered by the association itself and describes the organizers' situation without citing concrete cases. Even so, it records a perception that matters to you if you make a living on stage: visual polish no longer serves as proof of experience.

What the public speaking references have in common

Read together, these statements point to the same place. Anderson talks about your own voice, Duarte about empathy, Gallo about lived experience, Treasure about authenticity and Brené about courage. Duarte, Gallo and Brené accept AI as support in preparation, and all of them put the speaker's story, voice and presence at the center of the talk. In the third article of this series, we show that the official TEDx guidelines follow the same line, with very concrete requests about language, rehearsal and reading on stage.

If you are going to open the next convention, one simple question remains: if these people watched your talk, would they recognize your voice in it?