Episode Summary
In this episode of the Pursuit of Knowledge podcast, host Manav sits down with Ken Babcock, co-founder and CEO of Tango. Tango is a rapidly growing AI company transforming how organizations create and utilize process documentation and workflow automations. Ken shares the story behind dropping out of Harvard Business School to build Tango, which now serves over 2 million users and is trusted by major companies like Walmart, Nike, SpaceX, Amazon, and Slack.
Ken dives into the evolution of Tango, which started as a browser extension for generating step-by-step guides of business processes and is now moving into automating workflows with AI-driven agents. The conversation explores how Tango addresses the pain points of traditional video-based training (like Loom), why structured data is essential for AI success, and how agentic workflows will shape the future of work by taking over repetitive administrative tasks.
Ken also touches on lessons learned from his time at Uber during its hyper-growth years, shares his perspective on the role of AI agents in modern businesses, and offers practical advice for founders about fundraising, co-founder relationships, and effective delegation. The episode wraps up with Ken’s personal insights on work-life balance, hiring for startups, and his favorite book recommendations.
Listeners get an inside look at the future of software, where AI agents will streamline work, free up human creativity, and fundamentally change our relationship with technology.
Transcript
Manav [00:00:00]:
Welcome to Pursuit of Knowledge podcast. My guest today is Ken Babcock. He's a co founder and CEO of Tango. Tango is an AI company that is disrupting the traditional software training model. Here's the kicker. Ken and his co founders built Tango while dropping out of Harvard Business School. Since then, Tango has raised over $24 million. Ken actually gave me a demo before we shot this podcast and I was qu shocked how useful it was, especially for a lot of delegation tasks. Currently the platform is trusted by over 2 million users and it's used internally in companies like Walmart, Nike, SpaceX, Amazon Salesforce and Slack. Ken, I'm really excited to have you on the show. How are you doing today?
Ken Babcock [00:00:53]:
Thanks for having me. I'm doing great. I'm in Los Angeles with you. Got to escape Chicago. Couldn't be happier.
Manav [00:00:59]:
So you dropped out of Harvard MBA to build Tango? I just want to dive right into it. Can you explain what Tango does?
Ken Babcock [00:01:06]:
Yeah. So we are a browser extension. I work on any chromium browser edge. We help people create documentation of their process in the flow of work. So it's as simple as clicking the Chrome extension, walking through your process and your core tools, and then outspits a beautiful step by step how to guide. And that's really where we started. We take that documentation and we're actually able to take over people's screens and walk them through exactly what they need to do. So it's shareable like a Google Doc. And now we're working on automations and agentic workflows. So we have all these breadcrumbs of what people do. How do we do it on their behalf?
Manav [00:01:43]:
So traditionally I think this problem was being solved by Loom. Right. Like you had to record a video and then explain in the video like how it works. And that was also like a slow method to show like a particular task because you had to watch the whole video. So is that what you're disrupting?
Ken Babcock [00:02:01]:
It's funny you say that Loom took off, I would say maybe three or four months before we really started Tango. So in a lot of ways, like we have grown in parallel with Loom. We did take a lot of inspiration from Loom and we actually have people on our team who worked at Loom. But when you look at video, it's the analogy I make is it's the full game film. It's everything we wanted to create, like the highlight reel. So what are the discrete actions that people are taking? How can we codify those into clear steps? And video falls Short, in a lot of ways, there's a performance art to video. A lot of people hesitate to even make videos because they're like, I really hate how I sound. And then on the other side, watching the video is just this constant, okay, I don't need the whole thing thing. I just need a little nugget. Like, how do I find where that nugget is? Are they sharing their screen? Are they doing it the right way? So we've basically come in and said we're going to take those processes and we're going to structure them.
Manav [00:02:56]:
That's a good point because I notice when someone sends me a loom video, I have to re watch it four times. And that could be something wrong with my attention span or it could be something about the way that person is explaining me. But with your tool, it's so clear that you cannot. Can you not follow it?
Ken Babcock [00:03:15]:
Yeah, absolutely. And with our Guide Me product, as soon as you open that workflow and you click Guide Me, Tango starts highlighting exactly what you need to do. You're not even looking at a reference document. It's guiding you in the flow of work as well.
Manav [00:03:28]:
I want to know what was the inception of the idea? Because it seems like it has evolved from where it started. So what was the original idea? And why did you decide to not complete your graduation but do it instead?
Ken Babcock [00:03:40]:
Yeah, so the initial insight, it sounds silly even calling it an insight because it's pretty obvious, or at least it's like right there in plain sight, is that so many organizations, so many roles rely on shadowing to upskill up level onboard new hires. And anyone who's been through that knows you're a little timid, you don't want to ask a lot of questions. You're watching somebody's screen, that person's going at their speed. You can't really follow. Maybe they're not even the subject matter expert. Maybe it's just like someone who's available approximate. And so what we said was like, is there a way that we can do this asynchronously? And now this was like probably end of 2019 when we first started working on the idea. What happened with COVID is you had organizations that all of a sudden went remote distributed. A lot of them laid off 20 to 30% of their workforce. So a lot of knowledge was leaving the room and people were realizing we don't have any documentation, nobody knows how to use our tools. And so that level of urgency that we were hearing from customers, that was for us, that was just a Cue the timing is now. We need to go build this thing now. And that was really the push to drop out of HBS was just the urgency with which our customers were talking.
Manav [00:04:54]:
Interesting. Of course, AI is the hot topic right now. And what's the role of AI? How is AI playing a role in building your product?
Ken Babcock [00:05:03]:
It's funny, we didn't really call it AI when we first raised our seed round, but we had a pitch deck and the pitch deck had this like crawl walk, run versions of Tango. And Crawl was where we were going to start. We're going to say, hey, we're going to replace this very acute pain. Point around documentation walk was, hey, we're actually going to guide people through the workflows. So we're not just creating the documentation, we're actually taking over their screen. That's more traditional digital adoption. And then the run side was, we're going to automate these things. We're just going to do it for people. Once they map it, we can then do it. We didn't call it AI, we didn't call it agents. It was 2020. No one was talking in that way. Yeah. And so what we have now is like the advancements in technology are meeting a product that, like you said, has 2 million users. We have millions of workflows that have been created. And so that structured data set is going to enable us to do a lot of really powerful things with agents.
Manav [00:05:58]:
Interesting. You told me something, you said 80% of enterprise data is unstructured. Can you explain?
Ken Babcock [00:06:06]:
Yeah, And I think everybody can relate to this, but when you think about just the data points that exist within your organization, some of it lives in emails, some of it lives in Slack, some of it lives in, like a call recorder, some of it lives in your CRM. And then all of a. All those data points are maybe incomplete or lacking context or they're. They're in the call recorder, but they need to be in the CRM. Right. And so they're misplaced. And so when we talk about unstructured data, and I think that 80% is a stat from Slack, actually, when we talk about that 80% of unstructured data, it's, it lives in all these places. It's not where it's supposed to be. It lacks the context that's necessary for someone to look at it and be like, okay, I understand what this means, or I understand the decision I need to make based on this. And so one way to look at Tango is we're providing a vehicle for people to structure their Data. Here's what I do, here's what I go through. Great. We're going to put it into, step by step, very clear, with screenshots, automate the descriptions, automate everything. And so we're creating a structured data set out of it. The reason that's so powerful is when you talk about AI, there are companies like Glean that are trying to solve this where they say, okay, we acknowledge that there's a lot of unstructured data. Now we're going to try to be like making sense of it for you. But the reality with that is it's garbage in, garbage out. So if your data is highly unstructured and difficult to make sense of, the AI products are going to have a hard time. And so we went further upstream and we basically said, let's create this product that forces people to structure their data around how they use tools. And so for us, that documentation layer, which will continue with our agentic product, that is such a competitive advantage because we're solving a pain point, but at the same time we're getting structured data back such that our AI can act on it.
Manav [00:07:58]:
I read somewhere that 65% of sales professionals tasks involve admin work. What does that mean? So if I'm a sales rep representative, I should be 100% focused on closing and talking to customers. But if I'm doing 65% of my admin work, then that, that's so inefficient. Is that a pain point you guys are solving?
Ken Babcock [00:08:19]:
It's a pain point we will be solving with our new agentic product. And you know that 65%, it sounds crazy, right? Two thirds of your day is admin work, but really think about that. You have, let's say a 30 minute call on your calendar with a customer. You probably spend 10 to 15 minutes prepping, reviewing old notes, maybe doing some research on the customer, maybe understanding, okay, what's my goal for this meeting? Maybe we had a previous meeting or maybe one of my other reps had a meeting. So there's 15 minutes of prep, conservatively 30 minutes of call. After the call, you review maybe your call recording or your handwritten notes or your typed notes or whatever it is, try to make sense of that. Then you put it in your CRM and then you make a decision. Is this an opportunity that's going to progress? Is this something that we're going to work? Okay, yes. All right, then let's move that opportunity. Let's say that takes 20 minutes and then you need to write the follow up. You need to basically say, hey, thank you for the time today. Here's what I heard from you. Boom, boom. I think this is what Tango can solve or insert software. Product can solve. I'd love to book more time. Can we bring in these people and do a demo? That probably takes another 10 to 15 minutes. So you're seeing now where this, like this administrative work is two thirds of what you're doing.
Manav [00:09:37]:
Yeah.
Ken Babcock [00:09:37]:
And that one third the selling, the thing that you're really good at.
Manav [00:09:41]:
Yeah.
Ken Babcock [00:09:41]:
And that you need to be exceptional at. You're only spending a third of your time on that. Now imagine if a product could actually take all that and do it on its own. You've now opened your calendar up. You're actually able to like move through sales cycles faster. And so there's all these downstream effects that I think we can unlock for teams.
Manav [00:09:57]:
That's so interesting. I was talking to a buddy at Equinox about. He's a. He's into commercial real estate. And before he used to cold call himself, and now he uses this AI tool called Terracotta. And this tool basically automatically leads voicemails. So all he's doing is scrolling Instagram reels and as soon as someone responds, he takes the call. So he does. So the Terracotta will leave a voicemail in his voice to all his outbound leads. And he uses AI notetakers to follow up with all these customers. So you're right, like at the end of the day. And of course, there's the AI doing cold calling as well. So at the end of the day, like, we're gonna have to only do the last leg of communication, which is just closing and disqualifying or qualifying the customer and just.
Ken Babcock [00:10:45]:
And building relationships. Exactly. You have more time to actually build those relationships. Which sounds counterintuitive because you're like, oh, AI agent, go handle all these tasks on my behalf. But the reality is, like, you can now think about some of the nuances. How you want to approach the meeting, how you want to show up, what you want your tone to be, what action you want to drive, all the relationship building stuff that happens outside the meeting. You can actually spend more time, like, strategizing.
Manav [00:11:09]:
Can you explain what agents are?
Ken Babcock [00:11:12]:
Yeah. So this term does get. Excuse my language, but like, bastardized. Everyone's, oh, we got an agent. What an agent really allows you to do, because you do have to train your agent. An agent is something that can autonomously reason through a workflow using some base of data that you provide it. So a lot of Agents will watch what you do and then they'll say, okay, do you want to turn this into a workflow? There's a lot of agents that you basically go into a flowchart builder and you build out the flow and you say, okay, here's what, here's one of our processes. The really powerful thing about agents though is unlikely robotic process automation. It doesn't need to be linear, so it doesn't need to be the flow builder that says, okay, end to end. Here are all the steps. Automate that. Because the challenge with RPA was always that those linear workflows aren't super complex. And so they weren't super high value, but they were redundant, they were low value stuff. It took it off people's plate. Great, whatever. Agents can actually handle workflows that kind of go like this. When they go back and forth and there's decision points, you might get kicked back to something. They're able to make a decision, they're able to understand, okay, what have people done historically? Based on the context that I'm ingesting and the data that I'm ingesting, what's the path that I should take? So it's pretty powerful if you think about all your data around all your workflows, something synthesizing that and making better decisions, and frankly, probably making better decisions than what like the average person would make.
Manav [00:12:45]:
AI is disrupting all the entry level jobs, the intern, all the tasks that you would give to an intern. Like, I think AI agents are going to be disrupting that. How are you guys using AI agents for? And you actually gave me a demo, so I want you to explain how are you using AI agents?
Ken Babcock [00:13:03]:
Yeah, we're using a bunch internally. I think some of the things that I described, we're doing that, we're using it on our CRM, we're using it on our call recorders. In terms of the product that we're building, the big differentiator for us is we are ingesting that workflow data in the flow of work. As easy as clicking around. We don't need API integrations, we don't need complex code to get in the back end of your tools to understand what you're doing, which a lot of agents do need. We're living at the UI layer. So for non technical users, that's a huge unlock because they're able to say, oh, wow, I can build an AI agent by just going through my process as I normally would. People are really intimidated by the idea of AI. There's some fear in there, but there's also some fear that am I going to be able to actually instrument this in my organization? So the approach that we took was like, how do we provide that bridge? How do we allow people to find agents approachable? And so we understood all this user behavior through our documentation, project production, that is going to remain. So that understanding of UX and how you invite people into the flow, how they create that stuff is now not just going to be a step by step guide. They're going to be able to create an agent that to me is really powerful, that anybody can use it out of the box.
Manav [00:14:18]:
You told me something off camera. You said the future is UI less. What does that mean?
Ken Babcock [00:14:24]:
Yes, the, the. Let's call this like the three to five year future. Yeah. I think SaaS is obviously a huge industry and to say that it's all going to go away, I would just be silly. But user interfaces were basically created for someone to interact with an underlying database. What is a CRM? It is a user interface to interact with all your customer touch points. Yep. And so there are buttons, there are folders, there are dropdowns, there are subfolders, There are all these things that help organize that database.
Manav [00:14:56]:
Like HubSpot or Salesforce.
Ken Babcock [00:14:58]:
Yeah, exactly. And that's true really of any software tool that you think of. Right. What is a Applicant Tracking System? It's a database of all your candidates and how you want to progress them is what the UI is. What is a hris? Human Resources Information System. That's a database of all your employees. Right. So when you start thinking about software as databases and the idea that we are going to expect people to interact with user interfaces, to be able to access the database when an agent can actually do it on their behalf. I don't need someone clicking buttons. And I'll go Back to the HRIs use case. Whenever I've worked at a company, I've maybe gone to the HRIS once or twice a quarter. And so for me to become like an expert in that, it's just silly. And so agents are going to take that away. I think what we're going to be doing in three to five years is actually talking to our machines more. We're not, we don't need to know this like complex lingo associated with this tool. We're not going to have to like follow workflows. We're just going to be speaking and typing to our machine about things that we need to get done. And it's just going to be triggers. Right. It'll be able to ingest all this data and make decisions based on that.
Manav [00:16:10]:
Yeah. I personally experimented a lot with these AI apps that you can code, like the wipe coding of you can make an app in an hour, that kind of thing. But then I noticed like debugging will take forever or, and then it becomes if the app is beautiful or the UI is easy enough to use for consumers or not. But I think we're seeing how AI is disrupting the whole app industry or like the software industry. Does that concern you?
Ken Babcock [00:16:38]:
If I was committed to my legacy business of internal tool documentation, training, guided walkthroughs, basically this underlying expectation that we want people to be good at using software, if I was extremely loyal to that, I'd be concerned. I think the thing that you have to think about as an entrepreneur is the business I'm building going to be a billion dollar business, particularly venture backed entrepreneurship. There's a lot of forms of entrepreneurship, but venture backed entrepreneurship, you have to ask yourself that question because any investor is asking that question first. And so for me it's our how to guides solving a pain point. Yes. Is software documentation critically lacking? Yes. In most organizations. However, do I believe that a billion dollar business is going to be built on how to guides in a world where like the future is UI less? Probably not. So I do think it's, it's going to disrupt software as we know it and you as a result as the entrepreneur have to disrupt your own business.
Manav [00:17:39]:
I would argue it might be worth a billion dollar business. Looking at the Loom acquisition, of course, yeah.
Ken Babcock [00:17:45]:
Which was a hair under a billion, but yeah, potentially that's the floor. Right. For venture backed, you have venture investors on the other side who are like, okay, we can value the business at this today, but then I need to Underwrite it at 7 to 10x or every firm has a different thing they underwrite against. But so you got to believe not only there's evaluation here, but there's also evaluation here and it's going to happen quickly. And it's yes, Loom was acquired for about a billion, but think it's. You even have to believe beyond that.
Manav [00:18:17]:
What's Ken Babcock's guide to fundraising?
Ken Babcock [00:18:21]:
Oh, fundraising is an art and a science. And I think Sequoia actually describes this as it's the numbers and the story. And so the science is what have you built? What are the metrics? Can we believe that there's something here? There's customer appetite, they're willing to use the product, they're going to Retain. Those are all the numbers. Yeah. And the story is really about where it goes, and you have to have a big vision. You have to have the vision that substantiates the underwriting of 7 to 10x. And I think sometimes people rely too heavily on the numbers. Hey, here's what I've built. Here are the results that I've achieved. That's important, but the story is just as important, and the numbers actually lend credibility to the story because it's like, hey, I believe this big, outlandish thing about the future, and guess what? I think I'm going to get there because I've been able to do this. And so they go hand in hand. And my advice always to founders is make sure you have the equal parts in your fundraising story.
Manav [00:19:22]:
Interesting. One of my favorite shows is Super Pumped, and when I've rewatched it, like, three times, and every time I'm watching that show, I'm like, who was inside Uber and experiencing that? And they must have been so lucky, experiencing the insane growth of Uber. Can you talk a little bit about, like, your journey at Uber?
Ken Babcock [00:19:44]:
Yeah. I was so lucky to be there. It was a crazy place in so many dimensions. And I actually thought the show did a pretty good job of kind of characterizing it. I think some stuff, like, the timeline was pretty condensed, and some people weren't maybe as theatrical as, like, a Hollywood actor would be. But regardless, like, they did a good job. It was an incredible learning opportunity for me. And I think there's something to be said for early in your career, seeing what success and seeing what winning looks like, because for us, every week was a record week. We'd go into all hands and we'd be like, new record week. It was almost like a joke. By the third or fourth year I was there, it was like, okay, yeah, every week's a record record week. And that distorts your expectations a little bit about building a company. What comes of that, how motivated people are, like, how loyal they are to the mission, what the downstream impact of all that growth is, how you manage that growth. Because growth is obviously what everybody wants, but it really stresses the system and it really pushes you. And so for us, the growth was like, there's demand out there that we can't meet. And so how do we just move faster and faster to be able to meet that demand? And obviously, we had a competitor in Lyft that was constantly at our heels. And so speed was like the name of the game. So you learned how to be fast. You learned how to Set up markets quickly. You learned how to meet customers, where they were. And so all of those things, which I probably didn't value as much in the moment. Now I'm like, holy cow, that was a once in a generation company.
Manav [00:21:17]:
You know what's crazy? I feel like no business is safe. Every business is fragile. I took a wayo coming here and to be honest with you, like, I prefer waymos now. I don't like taking an Uber.
Ken Babcock [00:21:29]:
Okay, that's fine.
Manav [00:21:30]:
How do you feel about that?
Ken Babcock [00:21:32]:
All my stock. Did you guys think about I cash out all my stock and Uber. So you can. Yeah, you can believe this.
Manav [00:21:36]:
Did you think of. I think they have some form of a collab. Right?
Ken Babcock [00:21:38]:
We, we had a. We had an autonomous vehicle department for a while. Extremely expensive to develop that hardware technology. And we weren't a hardware company.
Manav [00:21:46]:
There was a whole story with the Levandowski.
Ken Babcock [00:21:49]:
Yeah, that too. I think way. Yeah, there is a partnership. But to your point, like every business needs to be looking in the rear view mirror a little bit. A little bit. Not all. You don't want to spend all the time doing that. But one of the sort of sobering things to do is look at the Dow Jones Industrial average from like 1940 or 50 or whatever. Look at it like 80 years ago. Very few of those companies remain. Most companies go out of business at some point and it's usually because of some sort of resistance to change. I grew up in upstate New York. Prime example of that is Kodak. We're here in a podcast studio. There's millions of cameras around. The camera didn't go away. The company just didn't think fast enough and didn't stay ahead of the market. They were very loyal to their legacy business, which is. Can be a trap.
Manav [00:22:39]:
I like that you are very agentic Forward.
Ken Babcock [00:22:44]:
Yeah.
Manav [00:22:44]:
In your thinking.
Ken Babcock [00:22:45]:
And it's not just. It's not. You want to be at the cutting edge for a lot of reasons. And for us, building agents is about being at the bleeding edge. And maybe an undervalued side of that is also your team members want to be working on the bleeding edge. They want. Engineers want to work on new. I interviewed an engineer today and he was like, I just want to be working on the hardest problems. And the hardest problems are the ones that aren't solved yet. And that type of mentality I think is. It's also energizing.
Manav [00:23:15]:
I want to know how do you think about delegation and as a CEO, especially because you're getting bombarded with 10 different things every day. So how do you personally manage delegation and how hands off are you? Any strategies you use for delegation?
Ken Babcock [00:23:31]:
I'm not going to sit up here and claim that I'm an expert on it. I think when you're a founder, there is always a tendency to want to like really get in the weeds and really want to understand something because that's what you do in the beginning and that's arguably like what you should be doing for a long time. We're still doing it. And delegation has been this ebb and flow for me. There have been times where I've said I need to delegate more, I need to preserve my time and my energy. And then there are times where you're like, man, I delegated too much and I feel like I've lost touch with this side of the business. And so the best strategy I think that I've settled on, at least for now, it's like picking your flavor of the month. I'm going to spend a lot of time over here. That allows you to get in the weeds, it allows you to scratch that itch. But you're not overwhelming yourself with, oh, there's only 24 hours in a day. And whether that's like sales, marketing, product development, customers like pick your flavor and just go after it and just keep a rotation because you can't be everywhere all at once. And I think that also is just. I see a lot of founders who are, who quickly burn themselves out.
Manav [00:24:34]:
That's a great advice. I want to talk to you about meeting your ideal co founders. That's the hardest thing. That's the number one reason most businesses fail is because the co founder dynamic. That's why YC rarely invest in one person companies. They always go for two or three co founders. So how did you meet your co founders and what's your dynamic?
Ken Babcock [00:24:56]:
So I met Brian and Dan at Harvard Business School. But it was funny. We were connected either before that or mutual friends in some sort of way. So I think it was like the first day honestly of class, we got together just more, more socially, honestly. And I think the more we talked about like our aspirations and what we were hoping to get out of business school and like what we were hoping to do next. It was very clear that there was alignment. Now just because someone wants to do entrepreneurship doesn't mean that it's going to work out to be for that person to be your co founder. I always joked that there are a lot of entrepreneurs at hbs, people that were like, yeah, I want to do entrepreneurship, but I'm Going to go intern with McKinsey and you're like, okay, nice. Good hedge for us. We were all pretty committed. Dan had been working in venture capital. Brian had exclusively started companies. The last thing I did before business school was a founder in residence. And I realized I didn't want to build something in an incubator. I wanted to build something, something from scratch. And there was a lot of alignment there. And then I also think, like, point in time is really important for co founders. Right. I think lots of people have experienced this where, oh man, I really want to work with this person I worked with before. But they're at this company and they're doing really well. I don't know if they're ready to, like, commit to a co founder relationship. We were all in business school. We weren't really tied to anything. Obviously we were at school and we wanted to be there. That's why we went. But we were also at a similar point in time where we could all commit to the company.
Manav [00:26:23]:
Did you read that Harvard MBAs were having a tough time getting a job?
Ken Babcock [00:26:27]:
Nah, they're just picky.
Manav [00:26:29]:
They're just picky.
Ken Babcock [00:26:30]:
Yeah, they're just picky.
Manav [00:26:31]:
I want to read a few quotes and then get your, like, little feedback on it. AI agents are so this was said by the CEO of DeepMind. He said AI agents are the bridge between information and action. And I think what he was trying to say. But I would argue that AI agents right now are pretty unusable, most of them.
Ken Babcock [00:26:52]:
I totally agree with you. But both those things can be true, right? I think the way I reframe that quote is we've been largely relying on systems of record. That's what a CRM is. It's a system of record. It's a database, whatever you want to call it, right? That's information. Agents are going to allow us to have systems of action, meaning the databases are still there, but now people are going to be able to take action on that via agents and they don't have to do it. So I totally agree with that quote on the usability piece. I think what's really hard and where I see a lot of agents failing is they build thinking they need to be a black box. They need to be the agent that gets triggered and just go off into the distance in some background tab and execute the task and then come back and be like, all right, we did it. Here's the result. I think the reality of where organizations are on the AI adoption curve doesn't match that solution. I think where Organizations are, is they still want to have observability. I talked to someone high up at Stripe and they're like, yeah, we have tons of agents running stuff. And I was like, really? How is that working? Do you trust it? Do you? They're like, oh no. There's a human review in every discrete task.
Manav [00:28:04]:
Interesting.
Ken Babcock [00:28:05]:
Because there's still nuance that they need to make sure they get right. The example she gave was in Australia you can't have a tanning bed business.
Manav [00:28:13]:
Why?
Ken Babcock [00:28:14]:
It's just, it's a law. There's no tanning beds in Australia. And so what Stripe needs to do is they need to make sure merchants are not tanning bed businesses in Australia. But an agent maybe doesn't have that nuance yet. Wow. And so they still need that human review, which I thought that was a pretty powerful story. And so the unusability is basically like people are building for that three to five year vision right now, but you need to bring users along for the ride. You need to build the trust with them.
Manav [00:28:46]:
So that comes to my question. Is AI agents going to be making company leaner or do you disagree with that?
Ken Babcock [00:28:53]:
I think yeah, absolutely.
Manav [00:28:56]:
That's the end goal. Right. Every company is trying to reduce their cost on CapEx.
Ken Babcock [00:29:00]:
Yes. I think it's also that they don't want, they don't want to have to handle administrative tasks. They want the people that they have to be focused on more creative, more strategic pursuits. So do I think it's something where all of a sudden like companies start laying people off? No, I think it's going to be more. We don't need that role right now. We actually have an agent doing that. So it's headcount deferral or headcount avoidance and it's not headcount reduction because all of these things in order to be successful are going to be augmenting the human that's already there.
Manav [00:29:33]:
Interesting. Okay. This is said by Sam Altman. He said eventually AI agents will be doing what junior employees do today and better. I think we both agree.
Ken Babcock [00:29:44]:
Or more. Yeah, yeah.
Manav [00:29:46]:
I think this is self explanatory. AI will be the ultimate personal assistant. Always available, infinitely scalable and deeply personalized. This was said by the CEO of Alpha a bit.
Ken Babcock [00:29:58]:
Yeah.
Manav [00:29:58]:
How personalized can it be?
Ken Babcock [00:30:01]:
I do think we've been talking a lot about like workflow agents. Right. Software executing tasks on your machine. There are other things that AI agents are being used for, particularly around voice. And so I actually just listened to a Talk from the CEO of 11 Labs and what they're focused on now. So they're a voice AI company, and so they. They provide basically AI call center agents that are able to interact with customers via voice. And one thing that he said that stood out is, like, imperfection in voice is actually what's going to make people trust AI call center agents more.
Manav [00:30:42]:
Yeah.
Ken Babcock [00:30:43]:
So for the longest time, they said they were focused on the perfect voice, the perfect intonation, the perfect everything. Even as we're talking here, live in person. Like, I've flubbed some words now. That's not an avatar. Right. And so that. That's how they're actually building it, is how do you accommodate some of these just like human elements that make people feel like, oh, wow, this is actually. This is. That sounds like me. Sounds like me.
Manav [00:31:08]:
Yeah, Yeah. I have another friend who is doing the same thing, AI cold calling. And he said the latency has become so much better in the last two years. But if. Let's say you asked me a question and I respond really fast, that can also, like, trigger some alarms.
Ken Babcock [00:31:24]:
Absolutely. There's a. I forget what the concept is called when you're ever on a loading screen, particularly think about a banking app or a financial app. If you're on a loading screen and it feels like it's doing something in the background, the phenomenon is that you actually trust that app more than if it was just instant. And it's a UX principle that I know our design team has worked on too, where it's like, give people a little bit of a lag. Not enough for it to be inconvenient, but enough. They could say, oh, wow, something's happening back there. I trust it.
Manav [00:31:56]:
Yeah. Another thing, I'm building a mobile app right now. One thing I learned is, like, the. The initial onboarding flow when they ask you a lot of questions.
Ken Babcock [00:32:03]:
Yep.
Manav [00:32:04]:
So the longer the onboarding flow is, the more people feel invested into your app.
Ken Babcock [00:32:10]:
Absolutely.
Manav [00:32:11]:
And if it's only two questions, then I might just close the app and forget about it. But if I've answered like, 15 questions about myself, you know how they ask you, like, all these questions like, how do you feel?
Ken Babcock [00:32:21]:
Like, how do you feel right now?
Manav [00:32:24]:
Yeah, yeah, it's quite interesting. Cool. Let's go to the next quote. This was by Bill Gates. He said, agents are not only going to change how everyone interacts with computers, they're also going to upend the software industry, bringing out the biggest revolution in computing since we went from typing commands to tapping on icons.
Ken Babcock [00:32:48]:
This is the UI list future that's upending software.
Manav [00:32:53]:
So I think a lot of people are thinking, how can I make money? How can I take advantage of this, this revolution that's happening right now? Let's make some predictions.
Ken Babcock [00:33:02]:
Okay. In terms of how we can make.
Manav [00:33:05]:
Money, not how we can make money, but where are we going five years down the road, like 20, 30, where are we going?
Ken Babcock [00:33:13]:
Yeah, I hope, and maybe this isn't money making, but I hope we're spending less time sitting at computers, trying to fight them to do things for us. I think a computer is going to hopefully be something that's autonomous and operating on your behalf. You're sitting here talking to me. Your computer might be actually performing tasks for you right now. That's where I hope we get to the Jarvis future. Yeah. And it just, I think we'll look back on, similar to how architects used to sit at these like huge drafting boards and draw things by hand. We're going to look back at like the modern workplace, the contemporary workplace right now and be like, wow, everyone is just sitting at computers all day, like bad for your health, bad for your eyes. What could they possibly have been accomplishing? Whereas instead, like maybe my desk is okay, I've got my machine over here that does a lot of tasks on my behalf and I'm able to do something else.
Manav [00:34:08]:
I love that. And especially with the GPT4O image generation model that came out recently, the capabilities are endless. You can do so much. I was listening to another podcast and they said AI is amazing, it's powerful. But what's a new idea that came out that was like so revolutionary? Nothing. And it was interesting because there hasn't been any mind bending idea that came out in the last three years that was new. That's something we already didn't know. I think it's just making us all more productive and whatever, more creative, I guess. But I want to know how you use personally. What AI tools are you using as a CEO of a really big company and. Yeah, except Tango, of course.
Ken Babcock [00:34:55]:
Yeah, I use, I'll give you one, one quick example and this is very much how I use it personally and not necessarily for work, but with the image generator that you mentioned. One of the cool ideas that I was inspired by was actually creating like a coloring book for your kids. So I have a three year old and a one year old. One year old's not coloring anything yet. But for the three year old I can use a photo of him and create a coloring book that's centered around him and the things that he likes and likes to do. And all of a sudden you've got, like, 20 images that you can stitch together, go print it out and give a custom coloring book to your kid where they're, like, looking at themselves playing with a dinosaur or looking at themselves on a scooter. And there's some really cool things like that too, where you're just able to.
Manav [00:35:43]:
That's incredible.
Ken Babcock [00:35:44]:
Yeah. You're able to do things that are more creative.
Manav [00:35:46]:
Hmm. And what about using GPT or ChatGPT or do you use that a lot for your daily workflow?
Ken Babcock [00:35:53]:
Absolutely. Every single day? Yeah, every day. Because there's. Think about all the mediums, which text is coming at you and you're trying to make sense of that text or you're trying to generate text. A lot of my job is also being, like, internal communications at the company.
Manav [00:36:09]:
Cool. So the people watching the show, they must be like, hey, I really like Ken. I really want to work for Tango. I want to know what's your high level on hiring people and what kind of people do you look for when you hire?
Ken Babcock [00:36:22]:
Yeah. And this has changed over time, but I think the biggest thing, particularly with startups, because I don't want to claim that we've made it or anything. We're still very much a startup and that DNA is there. So I want people that are comfortable with ambiguity, show an ability to, like, change their mind, and are strong collaborators. And these all tie to, like, core values that we have within the company. We are going to be working hard together. We need to check our assumptions. And to me, there has to be an appetite for risk too, because a lot of what we're doing, you're charting a new path and there's a lot of unknowns.
Manav [00:36:58]:
I love that. Last question. How do you unwind?
Ken Babcock [00:37:02]:
How do I unwind?
Manav [00:37:03]:
Because I feel like people in America, work people, we don't. We don't really have, like, things to do on the side. But you have a family now, so I guess you're pretty occupied with that.
Ken Babcock [00:37:12]:
Yeah, I'm pretty. I. My time is pretty much accounted for, but I've always played a lot of sports ever since I grew up. I continue to play racket sports and things like that. And that, to me is always. Exercising is crucial to making sure that you can also, like, show up in a good headspace at work, do a lot of that.
Manav [00:37:29]:
And any books you would recommend to other people like that you. That personally change your, like, way of thinking.
Ken Babcock [00:37:36]:
Yeah, for books. I try not to read too many business books.
Manav [00:37:40]:
Yeah.
Ken Babcock [00:37:40]:
Because it's just. You need an escape too. I'll give you two though. Lost City of the Monkey Gods is an amazing book. It's basically about this. It's like a documentary almost of this guy who's going into one of the few uncharted places in the world. And there's all this mythical lore about it too in the culture there. So that's a great one. That's one I've given to tons of people. And then one I just finished, it's about three years old now. It's called 4000 Weeks. It's by Oliver Berkman. It's basically just like 4000 Weeks is the average lifespan of a human. And it's just like how you use your time. But not only that, cause it's not a productivity book. It's actually. There's a lot of anti productivity things in there which I appreciate seeing that side too. It's. It's just giving you a different relationship with your time. Understanding how limited it is and understanding that like the point is not always to get to the next place with your time. It's actually to be present.
Manav [00:38:36]:
Time is the most valuable resource we have. That's so well put. 4,000 weeks.
Ken Babcock [00:38:41]:
It's a great book.
Manav [00:38:42]:
It's not long at all. Like time is so precious. How can people find you? And how can people learn more about Tango and actually use the product?
Ken Babcock [00:38:51]:
Yeah, look, we're. You can. We have a free plan. You can get started on Tango today. Yeah, we. We're in the Chrome store, the Edge store. You can just go download it. We're a browser extension, so that's really easy. If people want to get in touch with me, feel free to add me, follow me on LinkedIn or reach out directly. Kenango us.
Manav [00:39:09]:
Okay. Thank you for coming on the show, Ken.
Ken Babcock [00:39:12]:
Thank you. It was great.
Related episode
