Resources / Blog

Bunch CTO Charles Ahmadzadeh: Artificial Intelligence for Human Leadership

Join Infinia ML’s Ongoing Conversation about AI

Episode Highlights from Machine Meets World

This week’s guest is Charles Ahmadzadeh, CTO of the AI leadership coach app company Bunch. Highlights from our conversation include:

If we can deliver the value without AI, we would always go with that solution because it’s a lot, usually, simpler, people understand it better. So doing AI for the sake of it, I think, has never really been part of what we do. We just ended up doing AI to deliver the solution.”

“Something very important to keep in mind when you design any kind of AI is not only how you intended to use it for yourself, but how others who gain access to this AI can use it.”

“I’m a firm believer that technology works best when it disappears. And that’s what you see in the products that you actually mentioned like Netflix and Spotify. You don’t see the AI, you just see the content and most of it is going to be relevant for you.”


Watch the show above. You can also hear Machine Meets World as a podcast, join the email list, and contact the show.

Photo by Markus Spiske on Unsplash

Audio + Transcript

Charles Ahmadzadeh:
It’s a bit ironic, no? The fact that we use AI to try to help people to be better leaders.

James Kotecki:
This is Machine Meets World, Infinia ML’s ongoing conversation about artificial intelligence. I am James Kotecki and my guest today is Charles Ahmadzadeh, the CTO and co-founder of Bunch, which is self-described as an AI leadership coach, Charles, welcome to the show.

Charles Ahmadzadeh:
Hi James. Thanks so much for having me.

James Kotecki:
So an AI leadership coach in an app, I think what it does is it gives you bite-sized kind of nuggets of leadership wisdom personalized to you. Is the AI we’re talking about here kind of like a recommendation engine, similar to Netflix?

Charles Ahmadzadeh:
Yes. That’s most of what it’s about. When we looked at how people learn about leadership, most of the time people have off the shelf solutions like reading books or listening to a podcast, which never really is going to adjust to you. Or you have the much more expensive options, which is having a personal coach, which for most people is just out of reach. Our mission at Bunch to enable anyone to become a leader. And I think that’s what we’re trying to achieve with this AI leadership coach.

James Kotecki:
And so you did approach this from the angle of leadership first and then the AI came after that?

Charles Ahmadzadeh:
Yes. I think if I have to think about how to solve a problem, I would always try to solve it in the most simple way possible. If we can deliver the value without AI, we would always go with that solution because it’s a lot, usually simpler, people understand it better. So doing AI for the sake of it, I think has never really been part of what we do. We just ended up doing AI to deliver the solution.

James Kotecki:
I do notice that in the marketing on your website, you highlight the fact that this is an AI-powered app. Tell me about that decision from a marketing perspective, because you could have hidden that, right? Like, Netflix doesn’t say, “we’re an AI powered video recommendation system.” The AI is just in the background. Why did you choose to highlight the AI?

Charles Ahmadzadeh:
When we looked at the overall market on how people get leadership coaching today, most of it is going to be either marketplaces or direct person-to-person coaching. And one thing that we wanted to help people understand very quickly is you’re not going to be talking to a human, it’s not going to be the traditional old fashioned coaching. It’s going to be content that is curated and personalized for you. And what we saw is that using the AI aspect is one of the fastest ways for people to realize that, “oh, this is an AI thing so it’s not a human.” But on the other hand, I do agree that’s a tricky aspect because I’m a firm believer that technology works best when it disappears. And that’s what you see in the products that you actually mentioned like Netflix and Spotify, you don’t see the AI, you just see the content and most of it is going to be relevant for you. And I think long-term when we have a bit more traction or recognition, this is most likely a place where we’re going to end up as well.

James Kotecki:
So tell me about the feedback mechanism here. I’m on my phone, I get this little nugget, this little quote from a famous leader. Am I kind of liking it or not liking it and that’s telling the algorithm the kinds of things I should get in the future? Because if it’s just based on my personal preference, then maybe that’s not actually giving me what I need as a leader.

Charles Ahmadzadeh:
Yeah. A very good question. There’s two parts that go into the recommender. The first part is just picking up patterns of behaviors between users. So if we see that someone has been on the same track as you, then we can kind of predict what’s going to be the next challenges that come. The thing that makes the recommender even more relevant is the measurement that we do when you start using our app. We do a leadership style test, a bit like the personality test, but a lot more focused on how you lead people. And I think that drives a lot of the personalization that you get. For example, I’m a very big introvert, it’s going to adjust the recommendations that I get.

James Kotecki:
Is there going to be a component of this in the future where your system is able to take feedback data from the quote unquote “real world?” You read my emails, you read my Slack messages, you get feedback forms from the people that are reporting to me, et cetera.

Charles Ahmadzadeh:
Yes. It’s actually some things that we have already been experimenting with. For example, connecting to the calendar that lives on your phone. We can know if you have, for example, a meeting with a lot of external stakeholders and we can give you a leadership advice on how to manage that. And I think the other big advantage of being on a phone is that we can build those features while having the standards that we want ourselves for privacy. And I think what we can do with the phones today, they are very powerful — and even on the iOS devices, there’s the Core ML from Apple — is actually do the machine learning processing on the phone and keep the data on the phone so it doesn’t even reach our servers.

James Kotecki:
How do you think about AI ethics overall here? What’s the definition that you use when you’re trying to build out your service, your product in an ethical way?

Charles Ahmadzadeh:
The two aspects of it that are really important for us is the goal of AI, which can be summed up as how will it be used and how can it be used. And I think that’s something very important to keep in mind when you design any kind of AI is not only how you intended to use it for yourself, but how others who gain access to this AI can use it. And the other one is going to be fairness, like how is it applied to everyone? Does it discriminate or reinforce the discriminations that already exists? And I think we always try to look at these two components.

James Kotecki:
When you’re talking about leadership coaching, what kind of biases, what kind of discrimination are you most actively trying to work against?

Charles Ahmadzadeh:
Oh, there’s so many. There’s so many, but I think the most common one that are found in the workplace are going to be gender discrimination. That one takes a lot of work and people feel very differently. Something that we see very often in the users that are on the app is a lot of work around imposter syndrome. It really influences also how people get to leadership positions. And part of the problem that we have today, not only the diversity in tech, but also the diversity in leadership, I think plays a big part of it. And so we always try to find ways through the app to empower people and give them the opportunity to take leadership roles.

James Kotecki:
You know, with what you’re doing, giving nuggets of leadership wisdom, those nuggets are also attached to specific people, which the users of your app may or may not resonate with for all variety of historical, personal, political reasons. How do you take that into account?

Charles Ahmadzadeh:
Yeah, that’s a delicate one. The only answer to that is diversity. What is it’s the diversity that we have in our team? Being a distributed team, worldwide, we have people in many different continents. But also diversity in our user base. While we’re heavily focusing on the US market, we have a huge pool from emerging markets and specifically India and Nigeria. And I think that plays a big part in getting the honest feedback that we need about what we’re putting out there, keeping us accountable to not reinforcing white supremacy and all those things that are so wrong right now. But I think that the diversity, both in the people who make it and the people who use it is the only answer.

James Kotecki:
How do you monitor the AI that you’re using to ensure that it maintains the ethical standards that you have?

Charles Ahmadzadeh:
Very good question. We monitor that every week. We basically spend a lot of time looking at how people consume the content, what content works for them, what content doesn’t work for them. And in every single step in the application, people have the opportunity to get feedback about what was good? What was not good? What can we improve? And surprisingly enough, we thought very few people would actually give us feedback, but a lot of people actually give us feedback. And so the team is constantly getting that, and then we act on it right away when it comes.

James Kotecki:
And I suppose the fact that your users explicitly know that this is AI maybe helps them make better recommendations because if they thought it was a person they might treat it differently from a recommendation perspective.

Charles Ahmadzadeh:
Yeah, exactly. I think a lot of people see that they can actually influence how the system works by giving more feedback. That’s part of the investment when you use an AI product, you know that the more you interact with it, the more it’s going to adjust. And I think that that might be one of the factors that plays into people giving us so much feedback on the app.

James Kotecki:
If we think about the future of AI and work, and we think about the jobs that would potentially be the safest from AI, a lot of people would put leaders, managers in that category. We need people to at least manage this stuff, we still need humans to make the ultimate decision. But what you’re doing is using AI to make those leaders better humans, better leaders. How do you see that? How do you sort through that combination of these two factors?

Charles Ahmadzadeh:
It’s a bit ironic, no? The fact that we use AI to try to help people to be better leaders. I think from a psychological perspective it makes sense to use AI to help people learn about leadership because there is this golden measurement in psychology, which is you have your own perception and then you have a peer perception and then you have an objective measurement. And the role that the AI plays here is the objective measurement. But I think if you look at the future of work, the trend that we see more and more is that people want to be empowered to be their own leaders and do the change that they want to bring to the world. So I think there’s not a place where we would replace the leaders, but rather instead of having a very few amount of leaders, enabling anyone to be their own leaders.

James Kotecki:
Charles Ahmadzadeh, CTO of Bunch. Thanks so much for being on Machine Meets World.

Charles Ahmadzadeh:
Thank you so much, James.

James Kotecki:
And thank you so much for listening and or watching. You can always email the show with your tips and suggestions, it’s MMW@InfiniaML.com. Please like us, share us, give the algorithms what they want. I am James Kotecki and that is what happens when Machine Meets World.