Arianna Huffington thinks AI could be a GPS for the soul

ARIANNA HUFFINGTON: Two years into building the Huffington Post. I was the divorced mother of my two daughters. I had bought into the collective delusion that in order to be a super founder and super mom, I didn’t have the luxury to take care of myself. So I literally ended up collapsing, hitting my head on my desk, breaking my cheekbone, and that was the beginning of this journey into understanding the epidemic of burnout in our culture, in our global culture.

RANA EL KALIOUBY: Arianna Huffington, founder of the Huffington Post, knew she needed to make some serious personal changes. But she wanted to do more. Now that she knew the dangers of burnout culture, Arianna wanted to address it head-on.

HUFFINGTON: I first worked to raise awareness around it. At the Huffington Post, I launched a dedicated sleep section in 2007.

I wrote two books on the subject, Thrive and the Sleep Revolution, and by 2016 I realized I didn’t just want to raise awareness, I wanted to help people change behavior, and that required leaving the Huffington Post and launching a behavior change technology company.

EL KALIOUBY: Arianna then founded Thrive Global, a company using technology to promote better health through behavior changes.

And now in the age of AI, Arianna partnered with Sam Altman to launch Thrive AI Health. And they’re building a personalized AI health coach.

Pioneers of AI is celebrating our one-year anniversary this month, which is why I am so excited to share my conversation with Arianna. I’ve been so looking forward to this interview for months now! Arianna and I talk about how a hyper-personalized AI coach could unlock health and wellness for all of us, the importance of prioritizing self-care, and how AI could be the GPS for our souls.

I’m Rana el Kaliouby and this is Pioneers of AI, a podcast taking you behind-the-scenes of the AI revolution.

[THEME MUSIC]

Hi Arianna. Welcome to Pioneers of AI.

HUFFINGTON: Rana, it’s great to be with you.

EL KALIOUBY: So I have a confession to make. I’ve been a long time admirer of your work and it’s been a dream to bring you on the show, so I’m so excited for our conversation.

HUFFINGTON: Me too. And I love all the things we have in common. You are the divorced mother of two. We both went to Cambridge.

Copy LinkHow a fearless upbringing can make risk feel possible

EL KALIOUBY: Yes. I wanted to start there actually, ’cause we both went there as international students. Maybe share with our audience that whole story.

HUFFINGTON: Well, actually, I love that episode of Meditative Story because it explains so much about my life. It explains about my mother who was a force of nature, and when I saw a picture of Cambridge University — nothing like Greece, where my mother, divorced from my father, was bringing up my sister and me in a one-bedroom apartment with no money.

I just saw this picture of Cambridge and I went home and said to my mother, I want to go there. I told everybody who would listen, I want to go there. And everybody said, don’t be ridiculous. You don’t speak English. We have no money and it’s hard for English girls to go there. My mother said, let’s find out how you can go there.

And she found out I could take my general certificates of education at the British Council after I learned English, apply for a scholarship. But the best part that explains her is that she never made me feel that if I didn’t get in, it would be the end of the world. She always made me feel, if you don’t get in, there’ll be another adventure.

So her life was an adventure and she always felt that we’re somehow held by the universe. A little bit like Rumi’s poem, live life as though everything is rigged in your favor. And she made me very comfortable with failure. She used to say failure is not the opposite of success. It’s a stepping stone to success.

EL KALIOUBY: I love that. One of my takeaways hearing that conversation was, you gotta embrace the journey without attaching to outcomes. And that’s become a mantra in my family. Literally, all my kids, my nieces, my sisters. It’s very powerful.

HUFFINGTON: Yeah, that is so true because if you are so attached to outcomes, then first of all, you don’t take as many risks because outcomes are not guaranteed, and if we’re so attached to everything being perfect, we often only play it safe.

Copy LinkWhy embracing failure made a second act possible

EL KALIOUBY: Arianna says that it’s this mindset that helped her take the leap back in 2016 – to leave the Huffington Post and start a new health-focused company.

HUFFINGTON: And it was a very tough decision. And back to my mother, I don’t think I would’ve done it if she hadn’t made me comfortable with the possibility of failure because I was leaving a super successful media company with 850 journalists, a Pulitzer Prize, offices in 18 countries.

EL KALIOUBY: Amazing.

HUFFINGTON: To start again. Raise money, hire people. I’m just beyond grateful that I did it because I so love what I’m working on. The fact that I believe our health, especially in this age of AI, is at risk and anything we can do to actually give people the tools and the right mindset to go through the turbulent times is going to be so important.

I wanna ask you a question. It sounds like when you were at the Huffington Post, you had this kind of notion that to be a successful CEO and leader, you basically did not have time for self-care, but then you did it all over again at Thrive. I’m curious if you do it differently. I invest in a lot of early stage founders and, having kind of my own experience as a CEO, I used to wake up every night at 3:00 AM and send emails to the whole company and that was really stressful for myself. Obviously no sleep, but also for my entire team. And now I advise my founders to really prioritize self-care. So is there a path to success where you’re also taking care of yourself?

Well, I absolutely built the Huffington Post after I changed my habits. I was at the Huffington Post from 2005 when I launched it to 2016, and my collapse happened two years in when the Huffington Post was very young and very much a startup. All our success came after, and I definitely profoundly believed that the changes in my habits were transformational in achieving the success the Huffington Post had. And I see that all around me. It’s a little bit like if you look at athletes, recovery is part of peak performance.

EL KALIOUBY: Totally.

HUFFINGTON: There is no athlete who doesn’t recognize that. So giving yourself time to renew yourself, to replenish yourself, makes you operate at the top of your game, in touch with your intuition, with your creativity, with your best decision making.

Whenever I’ve made bad hiring decisions, I can go back and look at the fact that I was sleep deprived. I was exhausted. I wanted to just basically check this off my to-do list.

EL KALIOUBY: Right. My coach calls it like, you’re below the line. Your nervous system is basically freaking out, and then you kind of come at it out of fear and scarcity, which.

HUFFINGTON: Yes, exactly. I love that.

EL KALIOUBY: In a minute, Arianna and I dig into how her AI health coach works, and why behavior changes need to start small. Stay with us.

[AD BREAK]

Copy LinkWhy self-care is a performance advantage not a luxury

EL KALIOUBY: So let’s dig into AI. I’m a big believer that AI can help amplify and augment human abilities.

We need to take a human-centered approach to AI and it can help us be happier and healthier and thrive better. And you are on this journey with Thrive AI Health, and you’ve teamed up with OpenAI to build a personalized AI health coach. So first of all, let’s just step back — what is the problem in our health system and why are so many people unhealthy?

HUFFINGTON: I think the fundamental problem in our healthcare system is that we have ignored the fact that there are five daily habits, five daily behaviors that govern our health. We think that health is what happens when we go and see the doctor. But as we say at Thrive, health is also what happens between doctor visits, and these five behaviors that we work on are sleep, food, movement — we don’t even call it exercise. Millions of people in this country and around the world don’t move. They lead entirely sedentary lives — stress management and connection. These behaviors are infinitely more important than our genes. Our genes are about 7% of our health.

EL KALIOUBY: Fascinating. And the rest is lifestyle.

HUFFINGTON: The rest is lifestyle and of course medical care if you get sick. But these lifestyle habits are absolutely critical. So my passion and what Thrive is based on is elevating these habits to be part of science and medical care. In fact, professor Ashley at Stanford recently said that exercise is one of the most important medical interventions.

So I just published a post on my Instagram and my LinkedIn that is a carousel that says exercise is one of the most important medical interventions, sleep is one of the most important medical interventions, food is one of the most important medical interventions, stress management is one of the most important medical interventions.

Connection is one of the most important medical interventions, and I say that, Rana, because they’re not warm and fuzzy wellness.

EL KALIOUBY: Mm-hmm. It’s science.

HUFFINGTON: Hardcore. Like if you are perpetually sleep deprived or if you eat ultra processed foods and junk, if you don’t move, you are going to be unhealthy no matter how many drugs you take.

Copy LinkThe five daily habits that shape health more than we think

EL KALIOUBY: Now it seems like these five things are very actionable and easy to implement. Why is it tough for people? And where does AI have a role to play?

HUFFINGTON: Yes. Well, behavior change is very hard. It’s not easy at all, but it’s doable with the right behavior change methodology, and we’ve worked with great scientists, behavior change economists like Kevin Volpp at Wharton and BJ Fogg at Stanford to actually break it down into what we call Thrive micro steps. Tiny little incremental changes every day that cumulatively become healthier habits.

EL KALIOUBY: And they compound.

HUFFINGTON: They compound. Exactly. But the key is to start small, which is the opposite of how we think of change. New year resolutions.

EL KALIOUBY: No more chocolate.

HUFFINGTON: No more chocolate, no more sugar, in the gym every day for an hour, et cetera, et cetera. So that’s fundamental. And the reason why AI and what we’re doing with our AI health coach is so important is because of the AI superpower of hyper-personalization. Behavior change is not one size fits all. So when we onboard someone on the AI coach, we don’t just bring in whatever data they’re willing to give us — biometric data, lab data, medical data. We also do an extensive preference onboarding. Like what foods do you like? What poetry do you like? What piece of music do you like? So in moments of stress, we can nudge you with things that are going to help you center. And also our nutrition director is an expert at swaps.

Like if you love Cheetos, she’s not going to tell you to start eating avocado toast. She will tell you to switch Cheetos to Power Curls.

Copy LinkHow hyper-personalized AI can make healthy choices easier

EL KALIOUBY: Yeah. I think the key to this vision of personalized AI comes from this trifecta of having sensors, tons of data, and both this idea of predictive and also conversational AI. Is this how you see it? Does Thrive connect to, like, I wear a Whoop, I think you wear an Oura Ring.

HUFFINGTON: Definitely. I think conversational AI is key. The difference with a coach is that it’s two-way — the coach nudges you. He doesn’t just answer questions. Let’s say if the coach also has access to your calendar and knows that at 2:30 you pick up one of your children from school.

And then you have a 3:00 podcast interview with Arianna and it takes 10 minutes to walk your child to school, the coach can say, Rana, you have 10 minutes to do some squats or extend your walk with your daughter so you get more exercise time in. So your coach is very involved with you 24/7.

EL KALIOUBY: I’ve been thinking about how memory is a really important moat in AI. So the AI that really gets to know you and has memory of your data and experiences is gonna be more powerful than one that doesn’t. How do you think about memory in this world of personalized AI?

HUFFINGTON: Well, AI without memory is useless. Memory and the large context window — billions of data points, potentially trillions of data points — is foundational to an AI model. So I think behavior change is totally based on AI’s powers for personalization and its memory capabilities.

EL KALIOUBY: Yeah. One of the things I’ve been really fascinated by — human memory is not monolithic, right? Like, we put more weight on some experiences than others. We try to forget some memories. We weight some memories more. And I believe AI needs to have a similar kind of more thoughtful approach to memory. It’s not just your entire history of all the interactions, and I don’t know if you’ve thought about that for Thrive.

HUFFINGTON: Well, what happens with humans is we often suppress memories that are painful or traumatic, but AI can help us uncover them and deal with them. One of our sayings is, express whatever it is that bothers you. First of all, you have to acknowledge it. Often people bury these memories so deep they don’t even acknowledge how painful they were, and they’re just there, festering in the darkness.

So the philosophy, the Delphi admonition of know thyself.

And not hiding your shadow, as Carl Jung would call it from yourself, is very important to a really full life. So if you acknowledge a problem, a grievance, pain, the next thing is to express it to yourself or someone you trust.

And the next is transcend it. And I think one of the problems with therapeutic culture is that people never transcend it.

Copy LinkWhy emotional healing requires more than replaying pain

EL KALIOUBY: What do you mean by transcend it?

HUFFINGTON: Like move beyond.

EL KALIOUBY: Yeah. Move.

HUFFINGTON: We’re not defined by our traumas. And so many people, including friends of mine, end up going to a weekly therapy session and regurgitating their hurts and at some point, after you have acknowledged and after you have expressed, can you please transcend. Otherwise, you become identified with your problem. Your trauma, your malady, even your medical diagnosis is not your identity.

EL KALIOUBY: Mm-hmm. And I suppose AI can help with that, right? Because it’s there 24/7 and it could potentially be a conduit to having some of these reflective conversations.

HUFFINGTON: Of course. There are also problems as you have seen recently, so I think it’s very important to keep working to get it right, and part of it is AI cannot be as sycophantic as it has been and simply reinforce whatever you are saying. Oh, Rana, oh my God, I know your ex-husband. Wow, what he did was terrible. That’s not exactly what a good therapist would do.

But AI has a tendency to tell Rana or Arianna what AI thinks we want to hear, and that’s really keeping us in a state of arrested development, which is incredibly painful.

And part of the therapeutic culture we’re living in explains why we have had a mental health awareness month or day for 76 years now. We have 1.2 million therapists, over 10,000 mental health apps, and the mental health crisis has never been worse.

EL KALIOUBY: Yeah. Why do you think so?

HUFFINGTON: Well, I think partly it’s because we have forgotten that we are also spiritual beings. And I don’t mean that in terms of any particular dogma or religion. I mean it in terms of who we are. I actually profoundly believe that Teilhard de Chardin, the Jesuit philosopher, was right when he said, we are spiritual beings having a human experience. But even if you don’t believe that we are spiritual beings having a human experience, we are definitely not purely human beings having a material experience.

EL KALIOUBY: That’s so true. Exactly.

HUFFINGTON: If you believe that, then it’s really important to find pathways to that center of peace, strength, resilience, love that is at the core of our being.

Copy LinkHow 60-second resets can interrupt stress and build connection

EL KALIOUBY: So you have one approach to doing that, which is your idea of a reset video, which I believe is integrated into the Thrive platform. First of all, what is a reset and why do you think it’s important? And then you’re gonna play yours and I’m gonna share mine.

HUFFINGTON: Yes. Wonderful. So resets are one of my favorite features on the Thrive platform because they are microscopic — they’re 60 seconds.

EL KALIOUBY: We all have 60 seconds, right?

HUFFINGTON: Exactly. Rana, I’ve never had anybody who said, I don’t have 60 seconds. Plenty of people who said I don’t have five minutes to meditate, but nobody says I don’t have 60 seconds.

They’re scientific. They’re based on the neuroscience that, while stress is unavoidable, cumulative stress is avoidable.

And it takes 60 to 90 seconds to move us from the sympathetic to the parasympathetic nervous system and to interrupt the stress cycle. All it takes is focusing on conscious breathing and images that remind us of what we love about our life, at the moment when we may not love our life at all. And also music, quotes, whatever you can — I mean, we have thousands upon thousands of resets people have created. Thrive has hundreds of preloaded resets.

They’re all incredibly different. As you’ll see from mine, there are pictures of my children. They’re not like the pictures you would put on Instagram. They’re pictures that mean something to you. They may be silly or funny, but they activate something in you.

EL KALIOUBY: Let’s play yours. You can watch Arianna’s Reset on the video version of this episode on YouTube or Spotify. What you’ll see are pictures of Arianna and her children, her grandchildren, and a sweet dachshund. There’s quotes like Rumi’s “Live life as if everything is rigged in your favor.” And at the bottom of the video there’s a gentle reminder – inhale, exhale, inhale, exhale.

It’s uplifting, sweet, and sentimental.

I love that. How often do you play your reset?

HUFFINGTON: Oh, many times a day. And you can create multiple resets with multiple pieces of music. Also at Thrive, for every team meeting we spin a wheel and one of us plays a reset.

EL KALIOUBY: Oh my God. I love that.

HUFFINGTON: Which is also a great way to connect with coworkers since we are all spread around the world. And now for people who’ve never met me or don’t know me at all, this gives them a little glimpse into me beyond my bio or my LinkedIn.

EL KALIOUBY: Yeah. It builds a real authentic connection. Okay, so I’ll share mine.

HUFFINGTON: Now you.

EL KALIOUBY: You can also watch mine on YouTube or Spotify. I included lots of photos of my children, my sisters, my parents, and some pictures and videos of sunsets and blue waters – my happy place. And my mission statement: “Create a smile and a spark in people’s eyes through the joy of learning and building.”. Yeah, what do you think?

HUFFINGTON: Oh, incredible. Absolutely. I want to play it for myself. That’s what is so great — playing each other’s resets also activates that peace and gratitude inside us.

EL KALIOUBY: Yeah, and it’s funny, when I was selecting what pictures and videos to add in there. You’re right, some of the pictures of my kids just bring a smile to my face. They’re a lot older now, but I’m like, ah.

HUFFINGTON: And I love the way you had your quote handwritten, so we get ideas also from each other. So when I do another reset, I’d like to copy that, so it’s a way to kind of stop and reconnect with something ancient and deep and strong inside us.

I highly recommend creating your own Reset – or at least taking 60 seconds out of your day, everyday, to breath and to reflect on what inspires you. It’s powerful what that can do.

EL KALIOUBY: After a short break, Arianna and I talk about guardrails, purpose and profit, and spirituality in the age of AI. Stay with us.

[AD BREAK]

Copy LinkWhat safe and mission-driven AI should look like

EL KALIOUBY: There’s been a very public and ongoing conversation around mental health and AI — AI therapists, AI friends. And obviously there’s a lot of potential good that can come out of this, but also a lot of concerns. So how do you create these guardrails around Thrive’s AI coach to ensure that people are using this in a safe way?

HUFFINGTON: I think we’re just at the beginning of this journey. There is a lot of work to be done starting with how we train the models, what the guardrails are. OpenAI is now launching something specific for teenagers. What is appropriate for children, appropriate for teenagers? We are just really beginning to look at all the problems and come up with solutions. And the one thing that’s different than with social media, we are going to react much faster. Because with social media, it took us a very long time to realize that we’re becoming addicted, that social media were tapping into the worst aspects of ourselves — our rage, our jealousy, our anger — and we now have an opportunity.

I’m not saying every company is going to do that, but we have an opportunity to realize that human beings are a mixture and they can tap into our better angels and still make money.

EL KALIOUBY: Yeah, I love that view. ’cause I really believe it — that you can create massive economic opportunity and also do good in the world. But I feel like that’s not a very popular view. But that’s what you’ve built both at Huffington Post and Thrive. So do you believe that you can drive both massive economic opportunity but also really be mission driven?

HUFFINGTON: Absolutely. And I believe increasingly so now because we are at such a pivotal moment, a seismic change with what AI is bringing to the world — geopolitical turmoil, polarization, but also good forces converging. People want to be empowered to take charge of their health. There’s a huge shift.

People no longer think that health is something to be left to their doctors. I love my doctor and doctors are incredibly important, don’t get me wrong, but if I went to my doctor, let’s say twice a year, and in between I didn’t sleep, I ate junk, I didn’t move, and I was stressed out — you could go to the best doctor in the world and it wouldn’t make you healthy. People are realizing that, and that’s a huge shift.

EL KALIOUBY: Yeah, which is amazing. You also talk about how AI can be the GPS of our soul. What do you mean by that?

HUFFINGTON: What I mean is that if AI knows what you resonate to, what music, what poetry, what quotes, pictures — then it can guide you closer to your soul, which is that center that I talked about. It can nudge you there at the right moment, and that’s an incredible power, used correctly.

EL KALIOUBY: Yeah, I think there’s a very, almost like an oxymoron — who’s gonna build that AI that’s gonna help you become more spiritual.

HUFFINGTON: Well, first of all, I think it all depends on what you train the model on. Our model is not trained on the whole of the web. If it’s trained on the whole of the web, you lose control. It can be anything. Our model is trained on Thrive’s behavior change methodology, on our micro steps, on our content, and on peer reviewed literature.

Plus on whatever whoever joins the coach wants to share. And again, that’s up to them what data they want to share with us. All the hallucinations and the problems happen when the model is trained on the whole of the web.

Copy LinkWhy humans should leave intelligence to AI and deepen wisdom

EL KALIOUBY: Right. Which is not relevant in your particular case to the application. So this is a question I ask of all my guests. What do you think it means to be human in the age of AI?

HUFFINGTON: What it means to be human in the age of AI is to recognize that for the first time, humans will never be more intelligent than AI. So AI will be more intelligent than your children, my children, my grandchildren, which is an incredible opportunity to realize that we are not defined by our IQ, that Descartes was wrong.

I think therefore I am is not what human beings are. So that means that we can actually develop the qualities that are uniquely human. Let AI be more intelligent than—

EL KALIOUBY: Yeah.

HUFFINGTON: And let humans be wiser than AI. We can give them intelligence. We’ll take wisdom, and for me, that’s going to be an incredible future. Wisdom and love. AI will never love. If we can have wisdom and love, that’s going to be an incredible age for humanity.

EL KALIOUBY: I love that actually. And I imagine wisdom is rooted in experience, right? So it’s something that AI, I guess, can’t get.

HUFFINGTON: Well, AI is not conscious, and Mustafa Suleyman wrote a great piece. Now let’s stop pretending that AI is conscious or let’s not create AI that appears semi-conscious. AI is a machine.

And it’s a tool, and I think it’s really important to make the distinction between what is human and what is a tool. And that’s why, Yuval Harari — seen as a doom sayer, but he has said what for me is the most profound statement about AI — which is, I’m paraphrasing, that for every dollar and every moment we invest in improving AI, we invest the same time and resources improving humanity. Then we’ll be okay. If we put all our bets on AI, that’s when we’re in trouble.

EL KALIOUBY: Yeah. I love that, Arianna. Leave the intelligence to the AI and focus on love and wisdom. What a great way to end our interview. Thank you so much for joining us on the show. It’s an honor.

HUFFINGTON: Thank you so much, Rana. Wonderful to be with you, and thank you so much for creating such a beautiful reset.

EL KALIOUBY: Arianna is one of my entrepreneurial heroes. I admire her personal journey – how she moved countries multiple times, how she lives life like the Rumi quote – that the universe is rigged in your favor. And I love how she brings a human-centric approach to AI that is rooted in love and thoughtfulness.

And of course, I can’t wait for all of us to have access to an AI health coach. I think this is one of the most powerful use cases of AI.

This wasn’t your typical AI conversation, but it’s an important one. Because to make AI that works for all of us and benefits our society, we need to center our very human experiences. We don’t want to just be technologically innovative, we want to be emotionally inspired as well.

Next week on the podcast, we’re switching gears and talking to President and co-founder of Harvey AI, Gabe Pereyra. They are the leaders in how AI is shaping legal work and changing the legal playbook. You don’t want to miss it.