Get Your Weekly Digest

Receive the best podcast summaries and insights delivered to your inbox every week.

We respect your privacy. Unsubscribe anytime.

TED Talks Daily

Can AI make us more human? A social psychologist and a business leader answer | Heidi Grant and Barry Cooper

with Heidi Grant, Barry Cooper

Published October 3, 2025
View Show Notes

About This Episode

Host Elise Hu introduces a conversation from the TED Intersections series in which social psychologist Heidi Grant and business leader Barry Cooper discuss how AI can support human learning, decision-making, and connection. They explore the importance of a growth mindset in a rapidly changing AI-driven workplace, how AI can transform feedback and training, and the emerging skill of prompt engineering. They also reflect on AI's role in personal habits, social media, and creative content, and where human empathy and shared experience will remain essential.

Topics Covered

Artificial intelligence Personal Development Growth mindset Decision making Critical thinking Tools for thought Workplace dynamics Social media and online platforms Curiosity Dopamine

Disclaimer: We provide independent summaries of podcasts and are not affiliated with or endorsed in any way by any podcast or creator. All podcast names and content are the property of their respective owners. The views and opinions expressed within the podcasts belong solely to the original hosts and guests and do not reflect the views or positions of Summapod.

Quick Takeaways

  • Continuous, "ravenous" learning and a growth mindset are crucial for thriving in a fast-changing, AI-driven career landscape.
  • AI can provide rich, objective, real-time feedback and simulations that significantly improve learning and development compared to traditional, sporadic human coaching.
  • Using AI as a shortcut to avoid doing the underlying thinking undermines expertise; using it as a coach to refine and challenge your own work strengthens it.
  • Good prompt engineering-asking precise, bias-aware questions-is central to getting reliable, non-hallucinated output from large language models.
  • AI can be a powerful partner in decision-making by surfacing hidden assumptions and biases that are hard for humans to recognize alone.
  • There will remain "moments that matter" where human empathy, emotional connection, and responsibility are indispensable, even if those humans are assisted by AI.
  • AI and personalization can either deepen isolating rabbit holes or be used to foster shared communities and "water cooler" spaces for like-minded people.
  • AI assistants can help people manage limited willpower and self-control by nudging healthier behaviors at the moments when humans are most vulnerable to distraction.
  • People will sometimes want AI-generated, hyper-personalized content, but the meaning of some art and media will continue to come from sharing it with other humans.
  • AI is likely to commoditize certain skills but also create demand for higher-value activities such as better decision-making, creative design, and sophisticated prompt engineering.

Podcast Notes

Introduction and context for the conversation

TED Talks Daily and TED Intersections framing

Elise Hu introduces TED Talks Daily as a show bringing new ideas to spark curiosity every day[2:27]
She explains this is the final episode of the third season of TED Intersections, an original series of unscripted conversations between speakers and experts[2:33]

Central question about AI and humanity

As AI continues to integrate into workplaces, Elise poses the question of how to ensure we do not lose what makes us human[2:44]
She sets up the discussion as being at the intersection of social psychology and business leadership[2:39]

Introduction of guests and topic

Elise introduces social psychologist Heidi Grant and NICE CX Division President Barry Cooper as the conversation partners[2:50]
She says they will discuss why the current era of AI is misunderstood and how working with AI instead of against it can help people grow inside and outside their careers[3:03]

Opening idea: AI helping us be our intended selves

AI revealing blind spots in human interactions

Heidi states that AI can help us be the self we are trying to be in interactions with others by pointing out our blind spots[3:11]
• She frames AI as capable of telling us what we might be missing in our behavior or communication with someone else

Career advice for young people entering the AI era

Barry's question about early-career workers

Barry asks what someone starting out in their career in the era of AI should be thinking about as they enter an entry-level job[3:49]

Heidi on continuous learning and growth mindset

Heidi says being a "ravenous, continuous learner" has been and will continue to be a secret to success in life[4:16]
• She notes this has been true even before AI "lit up" in a big way in recent years
Heidi distinguishes between the idea that everyone thinks they love to learn and the reality of loving to learn as an adult whose performance is constantly evaluated[4:37]
• She explains that adults face real consequences for success and failure, which makes genuine learning harder
She describes a common mindset where people focus on looking like they already know everything, becoming terrified of not appearing knowledgeable[4:51]
• This fear leads to wanting to stay safe and stick with what is familiar instead of taking learning risks
Heidi stresses that people must fight this tendency because the pace of change is not going to slow down[5:11]

How AI is transforming feedback and coaching in customer service

Barry describes traditional human-based feedback in call centers

Barry explains that his company works in customer service and uses AI in that space[5:14]
He says that historically, customer service agents on phones or chats had their performance reviewed by a coach or quality manager who listened to a small fraction of calls or emails[5:27]
• A manager might listen to 1 in 10 calls or review 1 in 50 emails and then give subjective feedback
Agents could argue that the coach happened to pick a bad example, claiming that other calls were great[5:48]

AI-driven performance monitoring and scoring

Barry describes software that uses AI to listen to every call and monitor every email an agent handles[5:50]
• The system scores employees on how well they perform, the behaviors they demonstrate, and the outcomes they generate
He says this allows organizations to understand areas for development beyond the opinion of a coach who might be in a bad mood[6:07]

AI-driven micro-training and simulations

Barry notes that there are clever ways of executing training using AI, including very specific AI-driven training for particular behaviors[6:22]
• One example is showing a demonstration of a call where a colleague gave a great example of the targeted behavior
He describes simulated customers: if an agent struggles in a certain situation, AI can simulate a customer to test that scenario[6:44]
• The AI can then train the employee by suggesting what they should have said or done differently in that situation
Barry calls this a great example of AI's potential in learning and development[6:57]

Heidi on why AI-enabled feedback is powerful for learning

Quality and objectivity of AI feedback

Heidi says Barry has touched on many important aspects of feedback and learning[7:04]
She highlights that feedback in human-to-human contexts is inherently subjective[7:13]
• She notes research suggesting that feedback says as much about the person giving it as about the recipient
She sees value in receiving rich feedback that is less grounded in a single person's subjective state[7:27]

Nudging behavior in the moment

Heidi emphasizes the importance of getting feedback at the moment when it can be used[7:37]
• She says nudging behavior is more effective when the nudge happens as the behavior is occurring
She notes that it is very difficult for human managers or co-workers to deliver real-time nudges, but AI can[7:43]

Safe practice of difficult behaviors with AI

Heidi is excited by AI's ability to create less threatening environments to practice challenging behaviors[8:01]
• She explains that humans inherently care about what others think, so any interaction with another person feels evaluative
• Even when people claim to have a growth mindset or create a "safe space," participants still worry about how they look
Heidi argues that AI coaching and simulations can feel less threatening and more freeing for trial and error[8:57]
• She says evaluation concerns may not vanish entirely but are reduced enough to encourage curiosity and exploration
She is particularly excited about AI creating safe spaces to practice difficult conversations, such as giving clients or bosses bad news[9:11]
• She notes that telling someone something they may not want to hear without demotivating them is very hard, and AI can help people practice this

Using AI wisely as a young professional: curiosity vs shortcuts

Stay curious and keep learning

Heidi advises young people entering the workplace not to lose their curiosity[9:33]
She encourages them to keep trying to "ravenously" learn as many things as they can[9:39]

Be thoughtful about AI-enabled shortcuts

Heidi warns that with AI, people need to be very thoughtful about shortcuts they may be tempted to take[9:43]
• Getting the job done in the moment with AI may be possible, but there will be a future price if you are not actually learning
She says people have to exercise willpower around either using AI just to get started or using it after they have done their own synthesis[9:58]
• One approach she suggests is using AI to help get started, but still personally doing the synthesis and pulling notes together
• Another approach is for the person to do the initial work, then have AI help refine it by pointing out what might be missing or how to communicate more effectively
Heidi recommends using AI as a coach that gives tips rather than asking it to rewrite work entirely[10:28]
• She acknowledges that this approach takes discipline but argues it will pay off when those who resisted shortcuts become true experts

Growth mindset, multiple careers, and changing skills in the age of AI

Barry on growth vs fixed mindset and career shifts

Barry agrees with Heidi and says she hit the nail on the head with the importance of a growth mindset rather than a fixed mindset[10:49]
• He frames growth mindset as recognizing that there is always room to grow and learn
Barry contrasts older career patterns with more recent ones: 70-80 years ago, many people had one profession for life[11:10]
• He says people would learn a profession and that would be their life's work
Over the last 20-30 years, he notes that people have typically had two or three professions as technology waves changed things[11:10]
Barry predicts that with AI, skills will change every couple of years, making constant relearning necessary[11:20]
• He says people will only get through this by maintaining a growth mindset and recognizing that what they learned two years ago may no longer be relevant

Prompt engineering and evolving AI skills

Barry compares AI to a high-performance car or amazing tool that not everyone can use equally well[11:42]
• He says being an expert prompt engineer is like being an amazing car racing driver who knows how to move the car very quickly
He cautions that the specific skills (like prompt engineering) will themselves change in a couple of years[12:01]

Concerns about AI use: learning loss, AI shaming, and accuracy

Leaders' reluctance and "AI shaming"

Heidi says she hears two major concerns from leaders reluctant to let their people use AI, including something she has heard called "AI shaming"[12:13]
• She describes this as discouraging employees from using AI
One concern is that people will not learn and will fail to build expertise if they rely too much on AI[12:31]

Accuracy, hallucinations, and evaluation of AI outputs

Heidi identifies the other major concern as accuracy, referencing headlines about AI hallucinations[12:37]
• She mentions AI making up law cases and citations for science articles that do not exist
She asks Barry how he thinks about these risks and how young people should evaluate whether they can trust AI outputs[12:57]

Barry on prompt engineering to reduce hallucinations

Barry answers the second part first, saying that asking the right questions of a conversational AI is key[13:22]
• He defines this as prompt engineering: asking questions in the right way to get the right answers
He warns that ambiguous questions open to multiple interpretations are more likely to produce hallucinations[13:27]
• He notes this is the state of technology today and suggests that in about a year it will likely be more self-correcting and better

Does AI make people lazy? Comparison to GPS

Barry addresses the concern that AI makes people lazy by comparing it to GPS in cars[13:42]
• He points out that people no longer carry paper maps and pull over every 10 miles, and in that sense GPS does make navigation easier
He argues that outsourcing navigation lets drivers focus on other things, such as listening to a podcast while driving[14:04]
He says AI will similarly change tasks we previously valued, like writing an essay, by commoditizing them and pushing humans to higher-value activities[14:09]
• He gives examples where people moved from sewing clothes to building machines that make clothes or working in fashion design, as production became commoditized
Barry predicts AI will commoditize tasks once highly valued but will elevate humans to higher-purpose, higher-value work if used correctly[14:59]

AI, confirmation bias, and decision-making support

Using prompts to avoid confirmation bias

Heidi enjoys thinking about how the right prompt can reduce the odds of a terrible answer[15:08]
She connects this to confirmation bias: if you ask AI to tell you why a specific claim is true (e.g., why Taylor Swift is the greatest songwriter), it will only give supporting evidence[15:19]
• If you additionally ask for people who disagree and why, you will get more balanced information
Heidi notes that this approach mirrors decision-making training, where people are taught to avoid biased questions[15:46]

AI surfacing hidden assumptions in decisions

Heidi recalls that traditional decision-making courses taught leaders techniques like engaging in critical thinking and surfacing assumptions[16:16]
• She points out that assumptions are almost inherently unconscious, making this very challenging to do unaided
She argues that AI can be a brilliant partner because you can ask it what assumptions you might be making[16:28]
• AI is likely to identify some assumptions, prompting the human to question their validity and seek information that may contradict them
Heidi says techniques taught for decades to improve decisions become much easier with AI augmenting the process[16:47]

Prompt engineering as a core future skill

Barry responds that what Heidi describes is prompt engineering: telling AI to validate assumptions is an example[17:02]
He notes that nobody had heard of prompt engineering two years ago, but he believes many entry-level jobs in the future will require it as a fundamental skill[17:12]
Barry imagines there will be degree courses (for at least a few years) on clever, smart prompt engineering to get the best out of AI[17:20]

Heidi on learning through building prompts instead of copying them

Heidi acknowledges that sharing massive "mega-prompts" is expedient, letting others reuse them quickly[17:34]
• She contrasts this with the deeper learning that comes from studying the process of prompt construction
To build a great mega-prompt, she says, you have to understand why each piece is necessary[17:55]
• Through this, people learn about concepts like confirmation bias and how question framing affects evidence retrieval
Heidi is excited about teaching better questioning and more thorough understanding of information through teaching prompting[18:18]
She also mentions that good prompting can help people think more rationally about risk, which is hard because of strong loss aversion impulses[18:28]
• She believes that with really good prompting, people can consider information in a more even-handed way
Heidi reiterates her excitement about the future of decision-making with AI, framing it as augmentation rather than AI making decisions for humans[18:40]

AI as an "overly confident intern" and the importance of humans in the loop

Bob Johanson's analogy of AI

Heidi quotes Bob Johanson, who likens AI in its current state to a really well-read but overly confident intern who is always with you[19:05]
• She says she loves this analogy and finds it a good way to think about AI

Human training and responsibility

Heidi notes that the intern (AI) is only as good as the training it has received and the person behind the wheel[19:14]
She stresses the importance of keeping a human in the loop[19:20]
Heidi asks Barry where he thinks humans will remain essential, and what areas we will always or for the foreseeable future want humans to play a role[19:27]

Human "moments that matter" and superhuman assistance from AI

Moments requiring human empathy and significance

Barry says different people will have different views, but there will always be "moments that matter" where human connection is needed[19:55]
• He lists emotional moments, situations requiring empathy, and events of massive significance that cannot go wrong as examples
He believes humans will be needed in those moments[20:15]

Humans leveraging AI to become superhuman in key interactions

Barry suggests that in many of those important moments, the human you interact with will be leveraging AI[20:15]
• He describes AI helping them be "superhuman" in solving your problem
He notes that their discussion has focused somewhat on business, but he is also interested in how AI will help people in their personal lives and human connections[20:33]

AI, personalization, and social media rabbit holes vs shared communities

Algorithms, rabbit holes, and loneliness

Barry brings up the familiar idea of "the algorithm" and the rabbit hole effect in social media[20:55]
• He says social media is so personalized that it can take people down very individualist, lonely places

Hope for AI to create shared communities

Barry expresses a strong hope that AI will be used to create shared communities and water cooler-type places of like-minded people[21:07]
• He imagines people getting together to share things either in the real world or in virtual spaces, instead of being isolated

Heidi's view on AI helping overcome human quirks and limitations

Heidi agrees in general and says she believes AI will help in those ways[21:25]
She says some of the most promising areas are where humans have quirks or limited capabilities that AI can augment[22:25]
• She starts to discuss smartphone distraction as an example where AI could help people be better humans

AI helping with willpower, habits, and being our best selves

Smartphone use, dopamine, and willpower limits

Heidi notes that phones are full of interesting things that provide dopamine hits but also dopamine deficits[24:42]
She points out that we often want people to put their phones down at night because research shows phone use before sleep is terrible for sleep quality[22:55]
She explains that nighttime is when we have the least willpower to do that, because willpower is a limited resource spent throughout the day[24:05]
• Every decision, impulse resisted, or restrained reaction (such as not firing off an angry email) depletes willpower
• She jokes that this depletion is why happy hour exists, because by the end of the day it all seems like a good idea

AI as a supportive partner for self-regulation

Heidi argues that AI can help us be the people we want to be in areas where we struggle, have blind spots, or are undermined by unconscious forces[24:55]
She suggests AI could act like a human partner who reminds you of your intentions, such as putting your phone down at a certain time[25:28]
• An AI system could propose an alternative activity, helping manage your day and offering tips like taking a break after two hours of staring at a screen
Heidi imagines AI knowing what specific five-minute activities you enjoy that can replenish you and "put gas back in the tank"[25:13]
She calls it exciting that AI could assist us precisely in moments where we need help the most in pursuing our goals[24:42]

AI-generated personalized art vs shared human culture

Debate with Barry's kids about AI and art personalization

Barry shares an argument he has with his kids about AI and personalization, especially in art[24:44]
He notes that we already have curated music playlists based on what we listen to and like[26:29]
He predicts that in the age of AI, music and even movies will be created specifically for an individual, not by an existing human artist[27:05]
• Such content would be "perfect" for that person in that moment, hitting all the dopamine they need
Barry calls such a personalized movie the ultimate rabbit hole: a lonely person watching a lonely, custom-made movie[26:25]
He says his kids argue that the value of a movie or song is partly that other people like it too and that others have watched or listened to it[26:33]
• For them, community and shared experience are part of why they enjoy media

Question of personalization scope: individual, group, or society

Barry wonders to what degree AI-generated content should be personalized: to the individual, to a group of friends, to a society, or to humanity[28:08]
Heidi calls this a great question and relates it to social-psychological theories of identity[27:07]
• She explains that identity is seen as a combination of who you are individually and your group identities
Heidi says people are unique individuals but also belong to meaningful groups along many dimensions[26:55]
She notes that in some cultures, group identities are more important than individual identity, and in others the reverse is true[27:42]
She predicts that people's desire for individual vs shared content will fluctuate even within a day[28:31]
• There will be moments when someone craves something uniquely for them and other moments when they want content they can share with others or that others share with them

Continuing importance of humans behind art

Heidi believes artists already are, and will continue to be, augmented by AI as a help in their creative work[28:45]
She also believes it will continue to matter to people that there is a human behind the art, music, prose, filmmaking, or design[28:31]
• She thinks that human authorship will likely always matter, even as AI augmentation increases
At the same time, she acknowledges there will be times when people just want something manufactured for them (like elevator music)[30:18]
Barry compares this to sometimes wanting a burger and other times wanting a beautiful meal at a nice restaurant[28:45]
• He hopes AI does not just become fast food for everyone, and that people will approach it thoughtfully like other life choices

Closing reflections and mutual appreciation

Heidi's closing thanks and outlook

Heidi thanks Barry and says the conversation was fun and that AI is a topic she cannot stop talking about[28:45]
She appreciates talking to someone she sees as at the forefront of what we are doing today with AI and who is also thinking deeply about what we will do with AI in the future[28:15]

Barry's closing thanks and expectations

Barry thanks Heidi and says it was good to speak to an expert in behavioral psychology[28:27]
He says the coming years will be interesting as the AI wave hits and affects people, and that Heidi has interesting work ahead[28:31]
Heidi replies that they will both have lots to do[28:45]

Elise Hu's outro and production credits

Elise notes that the conversation between Barry Cooper and Heidi Grant is part of TED's original series and invites listeners to watch it and others on TED.com[28:45]
She mentions that listeners can learn more about TED's curation at TED.com slash curation guidelines[29:00]
Elise lists members of the production and editing team and says she will be back tomorrow with a fresh idea[29:25]

Lessons Learned

Actionable insights and wisdom you can apply to your business, career, and personal life.

1

Treat AI as a coach and collaborator rather than a shortcut, doing your own thinking first and using AI to refine, challenge, and expand your work instead of replacing it.

Reflection Questions:

  • • Where in my current work am I letting AI do the heavy lifting instead of using it to sharpen my own ideas and reasoning?
  • • How could I redesign one recurring task this week so that I draft first and then ask AI for critique, alternatives, or missing angles?
  • • What specific boundaries will I set for myself around when I will and will not let AI rewrite or generate content for me?
2

Cultivate a growth mindset and continuous-learning habit, recognizing that skills can become outdated within a few years and that adaptability is now a core career competency.

Reflection Questions:

  • • Which of my core skills are more than two or three years old, and how confident am I that they are still current in an AI-enabled world?
  • • How might I build a regular learning routine (courses, practice, experimentation) that keeps me slightly uncomfortable and growing rather than staying in the familiar?
  • • What is one concrete learning goal I can set for the next 90 days that would make me more adaptable to future AI-driven changes in my field?
3

Use AI to improve the quality of your decisions by deliberately prompting it to surface hidden assumptions, opposing views, and potential biases like confirmation bias and loss aversion.

Reflection Questions:

  • • In an important decision I'm facing now, what assumptions am I aware of, and which might I ask an AI tool to help me uncover?
  • • How could I incorporate a step into my decision-making process where I explicitly ask AI for the strongest counterarguments to my current plan?
  • • What upcoming decision could I run through an AI-assisted "assumption check" to practice this way of thinking before the stakes are high?
4

Design your prompts carefully and specifically, since the questions you ask AI strongly shape the answers you get and can either reduce or amplify errors and hallucinations.

Reflection Questions:

  • • When I last used an AI tool, how precise and unambiguous were the prompts I gave it, and what impact did that have on the quality of the output?
  • • How might I rewrite one of my typical prompts to include constraints, desired perspectives, or explicit requests for pros and cons?
  • • What simple checklist (e.g., ask for alternatives, ask for counterarguments, define context) could I apply before I hit enter on an important prompt?
5

Leverage AI to support your self-regulation-nudging you toward better habits and timing-especially in moments when your willpower is weakest and you're most prone to distraction.

Reflection Questions:

  • • At what times of day am I most likely to make choices I later regret (like doom-scrolling or working through exhaustion)?
  • • How could I configure an AI assistant or digital system to gently interrupt me at those times and suggest healthier, restorative alternatives?
  • • What is one specific behavior (sleep, focus, breaks, phone use) that I want AI to help me manage differently over the next month?

Episode Summary - Notes by Kendall

Can AI make us more human? A social psychologist and a business leader answer | Heidi Grant and Barry Cooper
0:00 0:00