Some of the most interesting aspects to me currently involve getting better at the “connection to communication,” the stuff that makes conversation with a crowd feel as easy and clear as a conversation with a person.
The telephone totally revolutionized the way we could communicate with people all over the world. But then came email and took it to the next level. And then came text messaging. And then came video calls. And so on…What’s next? What’s just around the corner?
In this interview series, called ‘The Future Of Communication Technology’ we are interviewing leaders of tech or telecom companies who are helping to develop emerging communication technologies and the next generation of how we communicate and connect with each other.
As a part of this series, I had the pleasure of interviewing Andrew Konya.
Born and raised in Cleveland Ohio, Andrew started his career as a theoretical physicist at Kent State University applying machine learning and supercomputing to understand emergent phenomena in complex systems. After an existential crisis brought on by seeing friends suffer on either side of a conflict in the Middle East, he shifted his attention to human problems. Now he focuses on applying machine learning to engage and understand populations.
He has spent the past 8 years working on collective speech and helping the world’s largest institutions leverage it — from the United Nations and the US government, to leaders of movements and F500 companies. In 2014, he co-founded Remesh with the goal of using AI to turn a live audience into a conversational representative intelligence. As CEO, he now leads a team of over 100 people working on solving collective speech and applying it to the world.
Thank you so much for joining us in this interview series! Before we dive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started?
I’ve always been interested in how we can understand the world and each other better. When I was very young it was religion. Then when I got older it was math, and the way numbers exist in the same way everywhere — two plus two is four anywhere in the universe! That grew into a love of Physics, since Physics is about describing the universe through math, and enables you to make new discoveries, and establish new truths, through hypothesis and experimentation.
My specialty was emergent phenomena in complex systems, what some people call the butterfly wing that disturbs the air in Africa just enough to lead to a hurricane in the Carribean. How do you spot that little thing that makes all the difference? How do large numbers of atoms, molecules, or particles interact in seemingly simple ways give rise to complex and unexpected behaviors?
So here’s how that relates to what we do at Remesh. I had two friends from grad school who returned to their homes, which happened to be places next to each other that are periodically in pretty violent conflicts. One day we were on a three-way video call, just getting caught up, when these two places started attacking each other. My friends kept talking with each other about the conflict, and the historical antagonism between their sides, points both sides had…but they ended up seeing enough of each other’s viewpoint for them to agree, “I don’t want bad things to happen to you, and you don’t want bad things to happen to me.” Then one of them said, Andrew, you’re a nerd, why can’t you build some kind of Artificial Intelligence system that can represent the population of a whole country, so our countries’ AIs can talk it out like we did?
That’s what got me thinking about how to scale real-time conversations with hundreds or thousands of people in a way that someone could understand, and help the group come to some kind of collective understanding. In the middle of all that information of people’s opinions and histories, how they talk about themselves and say what they want, what is the emergent behavior that can bring them to better understanding? It’s a really interesting problem, since people and societies are a lot more complex than particles! With particles, you have standards like energy, location and direction that you can count on. People have emotions, choose things with imperfect information, make trade-offs, and all use language just a little bit differently. Yet you can help them come together.
But as I thought about it, which is another way of saying while I sweated over making it work for a couple of years, I became even more convinced it was a big, important problem with a potential for super high impact. Not just for conflict resolution; this is a way of understanding a population, fast. Where is that important? Teachers need to know about their classrooms. Politicians need to know what their voters want. Good companies want to serve their customers better. The dynamic is important everywhere. Particularly in an Age when the world is changing fast, and there’s less time between things like the creation of new products, the rise of protests and social movements, or the creation and execution of marketing and advertising campaigns.
We’ve made exceptional progress, but I’m happy to say I think I’ve found a problem I can work on for a long, long time. Modern physics has good systems of measurement. By comparison, what people have been using for understanding collective opinions are pretty primitive.
Can you share the most interesting story that happened to you since you began your career?
Almost every “Remesh,” what we call one of our large-scale conversations, is interesting. People are always interesting, and whether we’re working on conflict resolution, how people feel about returning to work after COVID, or their hopes for their economic life, they’re talking about some of the most vital things in their life. Really, it’s kind of a privilege to be part of it.
But let me give you two: We will never take credit in things like conflict resolution, because we go into that with exceptionally dedicated professionals who spend years taking personal risks to end killing. They deserve the credit. But last year, I got to watch as Remesh played a real role in ending a lethal conflict that was bedeviling a North African country. The government there was not strong, and the two sides had an age-old enmity. Seeing them agree on basic terms of things like ending violence, making the way for COVID aid, holding fair elections for a new government, and building out a first basis for understanding, was really gratifying.
The other one is a lot closer to home. We use Remesh in our office on a weekly basis, as a way of guiding how people are feeling about the work, our markets, things we’re doing well and what we could do better. As a leader, who still has to make big decisions, it’s an exceptional tool for keeping me aligned with our goals. Plus, in normal times it even helps us decide what we’ll do for the End of the Year party. It’s very cool that the technology I originally envisioned for peacekeeping is now helping me be a much more effective CEO myself. And, it’s super meta that Remesh tech is helping the organization that makes Remesh tech make Remesh tech even better. Never underestimate the power of a compounding feedback loop!
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
First, a little context. When you hear that startups are hard, believe them. If I’ve given the impression that this was a slam-dunk idea that the world immediately applauded, that’s definitely not the case. It has taken years. It is the work of a lot of super smart people, testing and experimenting with cutting edge tools. There’s Natural Language Processing, a kind of AI, just to handle the language understanding. There’s other AI to find the basis of common ground among people. There’s algorithms for getting the right information in front of a conversation moderator, who leads the Remesh — she can’t digest and make sense of 1,000 voices, she has to know the key elements, and how to move the conversation forward. And that’s before you get to making a product, finding the right market, securing financing to keep the lights on, managing people so everything is happening at the right time.
To get to where we are today, we’ve built five completely different versions of our platform, and a similar number of business models. Even more organizational models. There was a period, now thankfully long past, where I was living on peanut butter and jelly. I had to sell my last valuable possessions, my guitars, to give us an extra couple of weeks of running room while we closed on the additional capital we needed to keep going. I didn’t think twice about it, since we were trying to solve a problem for the world, but that doesn’t mean it was easy.
So you can see why the quote from Morris Wheeler, an entrepreneur and angel investor, matters to me. He says, “Be a cockroach.” Survive at all costs. We’re passionate about this idea, and the difference we can make. But to succeed, you first have to survive. To survive, you have to not die, you have to keep going. That comes at a cost, but it’s how you make a real impact.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
Just one? There are so many. Robin Selinger, my Ph.D. advisor, showed me how to create an environment of continuous learning, and pushing hard to get what you care about. Cliff Reynolds, who managed an Ohio innovation fund, was the first person who rejected granting money to Remesh — and taught me some brutal realities about building a business, eventually becoming my first investor, too. Charles Stack, our first equity investor, spent many hours teaching me how to be simple and clear in explaining Remesh and how we work. Jenny Fielding, who directed the Techstars program I eventually attended in New York, really taught me how a significant startup operates, combining a lot of empathy with a lot of radical candor in pushing me (and Remesh) to raise our game.
How have you used your success to bring goodness to the world?
I mentioned the conflict resolution stuff, which is obviously good. I think you can find a lot of goodness at a smaller scale, too.
Let’s just take the business context. On a micro level, using this in the office helps me treat people better, since it builds empathy and understand Remesh’s whole team better. The decisions are mine, but my reasons are clearer. If people know I’m listening and acting on that, there’s more trust. Moving up from there, higher trust environments organize better, act faster, and help people do more valuable work. That helps them find more meaning in their work.
Getting even larger scale, there are all kinds of alignment problems we can attack. There are political things, of course, like what policies voters really support, but think about all the places where leaders (or machines!) make decisions that impact the future of entire populations or communities. Just commercially, maybe $5 trillion of spend a year is informed by researchers trying to understand what people want. That’s equivalent to all of humanity working for 20 straight days. Improving that alone represents a huge impact towards helping the future we create be the future we want. Not to mention the potential to impact the other 345 days.
Can you tell us about the cutting edge communication tech that you are working on? How do you think that will help people?
I mentioned a few things earlier around NLP and AI. That field is changing fast, with new and more powerful methods showing up all the time. I’m lucky to have teams that like to learn and innovate.
Some of the most interesting aspects to me currently involve getting better at the “connection to communication,” the stuff that makes conversation with a crowd feel as easy and clear as a conversation with a person. From birth humans learn to use conversation as a tool to understand the world around them. By tapping into this, and enabling it to work not just with one person, but with hundreds, thousands, or millions, makes engaging and understanding a population accessible to the masses. Figuring out how to spot a seemingly incidental part of a massive conversation that could turn out to be a key to new and better understanding, that’s fascinating. My hunt for the butterfly wing, if you will, only among the far more complex domain of people.
How do you think this might change the world?
I strongly believe we can change the world for the better by making people feel better understood. Empathy is a critical dimension to success in many areas, and sometimes it’s hard won. We can make some of that easier. All in, I think this helps better align the direction of progress with the will of humanity. It makes the future we create more likely to be the one we all want.
And, getting back to my childhood passion, that’s also a means to self-understanding and truth.
Keeping “Black Mirror” in mind can you see any potential drawbacks about this technology that people should think more deeply about?
Remesh makes it way easier for a lot more people to deeply understand a population. When this understanding is used to create things that improve people’s lives, that is great. But if that understanding is used to manipulate people’s beliefs about what can make their lives better, that is super dangerous. Beyond that, Remesh can empower a community of people to have a voice and act as a single unit. And if that community has intentions of malice, that can be a bad thing.
But at the end of the day, Remesh is both a truth machine and an empathy machine. And I can’t think of a better way to fight manipulation and malice than more truth and empathy.
As ever, though, being good involves being vigilant. And a lot of brain cycles continue to get spent on ensuring Remesh is used for good. I think to assume we’ve achieved success on this front is to ensure failure.
Was there a “tipping point” that led you to this breakthrough? Can you tell us that story?
The successes in conflict resolution and political understanding were encouraging, as results. They also taught us a lot about some of the technical problems we’d have to solve to get to where we wanted. So let me spend a minute on the technical breakthrough.
The primary technical challenge of Remesh is being able to accurately represent a population’s response to a message or question we don’t know ahead of time; starting within nothing, and then gathering the data, analyzing it, and displaying that representation — all within about a minute, so live dialogue can happen.
This presents two seemingly opposite technical challenges: expansion and compression. On the one hand, because we only have a minute, we can’t gather all the data we want from the population. We just get a small sampling. So the first challenge is to take that small sampling and predict all the data we are missing. This is a problem some know as “compressed sensing.” You can think of it as trying to squeeze all the information you can out of the data you actually have. The first breakthrough came in figuring out how to do this really fast, within a wide range of scenarios, in a way where we could trust the results every time.
But this brings us to the second challenge. You now have even more data about what represents the population — you have as many responses as you have participants, and as many opinions about responses as you have participants squared! Yet, you need to convert all of that into something a human can digest in less than a minute to genuinely understand the population. It’s a problem of filtering and compression, turning a massive amount of data into a small amount of information that contains what matters most.
Physicists have to deal with this all the time. The best example is the large hadron collider. When it’s running, it is smashing particles together at the speed of light generating nearly 30 million events per second, or about 2 exabytes of data in a day. This is more than can even be stored, let alone processed. To solve this, brilliant physicists and mathematicians have figured out ways to filter the data and extract the ~0.004% most likely to contain the information that holds the signals they are looking for.
In a way, our problem is simpler — the amounts of data we have to generate and analyze are many orders of magnitude smaller. On the other hand, our problem is harder — the relationship between humans, language, and what represents them is much more complex and imprecise than the trajectories of particles and the models that describe them. But we’ve made great strides in bringing a quantitative approach to this seemingly qualitative problem.
These breakthroughs are enabling us to bring a new communication paradigm into the world — many-to-one conversations — and the beneficial impact on humanity will be profound.
What do you need to lead this technology to widespread adoption?
Without getting too heavily into the financials, we were very pleased with the revenue and growth numbers we showed for our last funding round. Since then, we’ve been accelerating at a rate that we are pretty happy about. Staying at this rate of adoption growth (knock wood), we’re on a path for global ubiquity before the end of this decade.
That is increasingly likely as we become better understood and appreciated. If everyone knew who was using Remesh, and the impact it was having on the world as they know it, I think that would certainly accelerate things quite a bit. But we take privacy seriously. We think in the long run trust beats hype.
There are definitely still a healthy inventory of interesting challenges for us, including finding new and better ways to get people involved in more conversations. One means to this is changing the economic models around participation. We already reward people for the data they provide in conversations, but things like representative intelligence sovereignty could make it much better . We also hope that people will become more aware of this kind of communications tool, and will demand of their leaders that they use it to listen better to them.
Leaders of all kinds are already seeing the benefits of this new method of communications. They see that it helps them do their jobs better. It’s going to be better for things like corporate candor and transparency. The organizations that use it will see follow on benefits as it serves as a signal to workers and consumers alike that their stakeholders get a voice. It’s going to be better for government performance. It’s already helped end conflicts, save lives, and shape the future — and we’re just getting started.
The pandemic has changed so many things about the way we behave. One of them of course, is how we work and how we communicate in our work. How do you think your innovation might be able to address the new needs that have arisen as a result of the pandemic?
By late spring 2020 we were already running Remeshes with workers about how they feel about going back to the office, what kinds of information they trust and what they’d like to see from leadership so they feel safe.
Looking forward, it’s safe to say we’re heading into new territory where work is concerned. It’s possible that some companies will hire people with the best skills, no matter where they are on the planet, and will try to build teams and company cultures virtually. Others may disperse to smaller offices with mixtures of home/office work. This could mean less of a need for office space, or a building boom, as social distancing puts fewer people than before in an office space. The way many of us work is in a volatile state of flux.
No matter which paths that organizations follow, it will be critical for them to know how people think and feel about an evolving workplace. Time to understand and adjust will be short, and flexible adjustment often critical. We can play a role there.
We might even become a selling point for corporate recruiters, since using Remesh signals a company prioritizes understanding and empowering their workers. “We care about the people who work here and give them a real voice in what we do” is a powerful selling point for the most in-demand talent.
That same uncertainty about the new workplace will be even more acute where consumers are concerned. Will there be a spending frenzy? Will they want to travel, go to big arena concerts, and celebrate? Or will they cocoon, seeking lots more information about safety and hygiene before they venture out. We’ve all got opinions, but nobody knows. And they’re going to want to find out, fast. Quarterly research on consumer sentiment could easily become weekly dialogs with consumers, more of a conversational relationship than distant research.
What are your “5 Things I Wish Someone Told Me Before I Started” and why? (Please share a story or example for each.)
- Writing is 10x higher leverage than speaking.
If you can explain something clearly in writing, it scales. There’s no need to keep repeating it. The bad news is, it’s really hard work to write something that will be understood equally by a lot of people. The good news is, when you write something clearly, you understand it better. I write things so I can figure out what I’m actually thinking.
2. Firing people is as important as it is painful.
Any boss reading this is nodding right now. Hiring, managing, and sometimes firing people is a deeply emotional experience for both sides. Sometimes you have to move people out because they are a drag on the team, or they’d grow better at a different organization. You owe it to your team to act.
3. Your gut is more correct than a physicist would think.
I’m a science-oriented person, but I believe there is such a thing as intuition: It’s a combination of the latent pattern understanding you build up over time in a specialized field, in other words, experience, and imagination of what else might be happening in your field. I’ve got good intuition where physics and startups are concerned; I’ve got none about being, say, a professional basketball coach.
4. Money matters.
Money is a store of potential energy. It is a means to accomplish more, faster. And managed right, it unlocks sustainability. Money is also a form of information. Where it is flowing, who has it, what people will trade for it, and what it’s accomplishing where — are all strong signals about the state of the world.
5. Trust is the key to efficiency.
When an organization has trust, it can minimize nonproductive governance frictions. There are fewer check-ins about a project’s status. There aren’t systems to check if people are making good on what they said they’d do. There’s more confidence that things will happen, which inspires people to take risks and to do more.
If you could inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger.
First off, thanks to those of you who’ve read this far. I hope you found it valuable. Be in touch if you want to learn more about Remesh.
The movement I’d like to inspire would be one that helps feel better connected: To themselves, to each other, to their leaders, and to the world. You can’t really extract one of those things from the other, and the breakdowns in connection are behind many of the world’s woes.
I think this is achievable, if we can solve our problems around empathy, alignment, and communication. We can invest in politics, work, even consumption, with more meaning.
I’d like to increase civic participation and corporate citizenship by creating “representative intelligence sovereignty,” by giving people real ownership and power over their data. What people do online, from surfing the Web to shopping and surveys is used to train models that represent and predict the actions of them and their peers. As an asset class, these models generate immense wealth for their owners. Today those owners are companies. I think they should be the people.
Giving people economic incentives in exposing, sharing and learning from their own data, with plenty of room for permissions and privacy, should be part of modern citizenship. That way people will have a stake in creating better understanding and more prosperity.
I’d like to see community dialog, with broader and more frequent conversations among all kinds of stakeholders. Much of the current tribalism and cynicism has to do with people feeling frustrated with the way they get and give information. We can make it better.
How can our readers further follow your work online?
Thank you so much for the time you spent doing this interview. This was very inspirational, and we wish you continued success.