Anti AI Slop Part One w/ Runbin Dong | Ep. 210

In this episode of Biz/Dev, we talk with Runbin Dong, founder of Scale Social AI, about what it looks like to take an AI-driven business to market right now, when the SaaS playbook is shifting and “authentic” is either real or it’s noise.

We get into building with intent, investing in yourself before anyone else does, taking full ownership when things get hard, and what it really means to start from scratch without pretending you’ve got it all figured out.

Runbin shares the mindset shifts founders need in this new landscape, the traps people fall into when they chase hype, and how to build something that actually earns attention, not just clicks.

Runbin on LinkedIn

Email Scale Social AI

___________________________________

Submit Your Questions to:

hello@thebigpixel.net

OR comment on our YouTube videos! - Big Pixel, LLC - YouTube

Our Hosts

David Baxter - CEO of Big Pixel

Gary Voigt - Creative Director at Big Pixel

The Podcast

David Baxter has been designing, building, and advising startups and businesses for over ten years. His passion, knowledge, and brutal honesty have helped dozens of companies get their start.

In Biz/Dev, David and award-winning Creative Director Gary Voigt talk about current events and how they affect the world of startups, entrepreneurship, software development, and culture.

Contact Us

hello@thebigpixel.net

919-275-0646

www.thebigpixel.net

FB | IG | LI | TW | TT : @bigpixelNC

Big Pixel

1772 Heritage Center Dr

Suite 201

Wake Forest, NC 27587

Music by: BLXRR

Transcript

[00:00:03] David: Hi everyone. Welcome to the Biz Dev Podcast Podcast about developing your business. I'm David Baxter, your host. Joined today per usual by Gary Voight. Hello, sir.

[00:00:11] Gary: Hey, what's up?

[00:00:12] David: More importantly, we are joined by the famous Runbin Dong of Scale Social AI. Hello? Good sir.

[00:00:20] Runbin: That was a lot, but David great to be here.

Not sure about Famous, but. 

[00:00:25] David: You are, man, I tell you, you are hot right now. I'm just saying it. I know you're gonna, you're very shy. I know this from meeting you a few times, but dude, I swear your name is everywhere. You got one of those, those startups that everybody's talking about and all excited and I'm gonna get into that.

I got, I, I have to be very nice here because, or nice to Gary. Last time we recorded, I skipped our fun intro section and he's very mad at me. So

before we get 

[00:00:50] Gary: Oh, no, no, no. I wasn't the one that was mad at you.

[00:00:53] David: You're right. Christie was mad at me. She's still mad at me. Christie's, our marketing person who's in charge of all things podcast, she's still mad at me, so I wanna ask, we start out with some intro questions, just get to know you a little bit better and see what makes you tick.

You ready?

[00:01:08] Runbin: Yeah.

[00:01:09] David: All right, here we go. What are you watching right now? Tv. Movie. What's your, what's? What's currently got you going?

[00:01:17] Runbin: Oh gosh, f just finished season two and I can't believe season three is coming in 2027.

That's a downside of like

fallout. 

[00:01:27] David: Fallout.

I watched the first season. I'm a big game fan, fan of the game. Is the second season as good or don't sound spoil it for me, but is this good? Yeah. Okay.

I'm 

[00:01:36] Gary: I'm a big fan of the effects, but I've never played the game, so 

[00:01:40] David: Dude that, yeah, it, it captures the game because the game is known for like, just over the top.

Weird

[00:01:47] Gary: I've, I know the imagery and the style behind it. 

[00:01:50] David: Very, very good. Alright. Are you a early riser or are you a midnight oil guy?

[00:01:58] Runbin: Wow. Good question. Used to be Midnight Oil.

[00:02:02] David: All right.

[00:02:03] Runbin: Now both.

[00:02:05] David: Both. Well, is that

[00:02:06] Gary: Well, when you got a startup, yeah.

[00:02:08] David: Yeah. Is that because you have to be? 

[00:02:10] Runbin: You don't sleep a lot. 

[00:02:12] David: when you as a startup founder, 'cause I I, I empathize with this, you're way ahead of me. But the, the vibe is still there. Is that because you're working or because you're now spending time with the family and stuff later because you're working so much?

Is that's what I find a lot of times.

[00:02:30] Runbin: Yeah, it's a little bit of both. In the earliest days, it's working constantly. It's

the fact that my business owners would text me around midnight, so I still have a habit of having my laptop right by the bed.

So when they text 

[00:02:42] Gary: Oh, 

[00:02:43] Runbin: like, I'll just pull it up and do some camera magic. But you know how it's,

[00:02:49] David: Nice.

[00:02:50] Runbin: now, like where we are right now, we have a really good team in place and so I try to be much more deliberate with like saving weekend times and evening times. And so we actually made it into a practice for our entire team where there's family time, like blocked out after 6:00 PM.

[00:03:06] David: I like

that. One of the things I did with Big Pixel early on was. I'm a big believer in family time and I would tell everybody, Hey, if I am working late, that doesn't mean you have to work late. And I had to specifically say that. 'cause when you're the leader you have to say that. And if I give you a weird slack message at eight or nine o'clock, that doesn't mean you have to respond until the next morning. And I would make that very, 'cause some people are working and if they can get to me, 'cause some people just like to work great. If you are ignoring me, that's fine too. So I gotta get it. Alright. Favorite hobby or pastime?

[00:03:41] Runbin: Gaming, big gamer.

And, 

[00:03:44] David: guy or you a PC guy

[00:03:46] Runbin: PC guy. and I get my console thing through steam, so I dunno

[00:03:51] David: You steam deck guy.

[00:03:53] Runbin: yep.

[00:03:53] David: We, I have found a lot of guys nowadays are buying steam decks so they can play in bed next to their wives. That's, I, I feel like that's cheating, but that I, that is, that is 

[00:04:04] Runbin: It's great.

[00:04:05] David: Definitely a thing. It's like, I'm here, honey.

What are you talking about? I went to bed the same time you did? Yeah. Yeah. That, that's absolutely a thing.

[00:04:13] Runbin: Indulges me and she's like, you gotta pull up your steam deck at some point. And it's great to have a partner who

[00:04:20] David: well, you gotta get the money's worth out of it. Right? Come on now. Alright, last one. Where do you think human robots. The assistance, the humanoid robots will be in five years.

[00:04:34] Runbin: Oh gosh. Have a, actually have a friend who's working in this area right now. Very specific tasks are absolutely possible. 

[00:04:44] David: You think they're gonna be in your house? Are they gonna be Elon's Dream where we all have one in five years?

[00:04:50] Runbin: yeah, I mean, just look at where LLMs have come in two years, right.

[00:04:56] David: for sure.

[00:04:56] Runbin: And so I think one of the biggest challenges right now with human humanoid robots is actually having a sufficient training data set, like sufficiently

large and diverse

[00:05:06] David: The world model. Mm-hmm.

[00:05:07] Runbin: Yep. To actually feed into like what the tasks are supposed to look like. That's the biggest limiting factor.

[00:05:15] David: well, I think and, and I, you're right, it's hard to discount how fast things have come, but at the same time, that problem is so large. Elon has proven himself to not be a reliable source of information in any of his business dealings. So when he says, I think his newest one is, he is gonna have more computing power, power and space by 2030 than all of the earth.

[00:05:38] Runbin: Mm-hmm.

[00:05:39] David: Alright, dude, you still don't have self-driving cars, I'm just saying. Not really, but I, I mean, I'm, I'm in, I'm intrigued by this. I don't know. I don't know. That's, it's alright, cool. So you're, you're optimistic. It sounds like you're optimistic that they'll be. All over the place. We'll have Rosie the robot in our house in five years. 

[00:05:56] Runbin: I've been thinking a lot about that actually because of the field that we, we work in. We are in the data field as well, right? And so if I wasn't working on scale social, I would've built some kind of a real world telegraph sort of device. Have you guys seen the order ring?

[00:06:12] David: I am 

[00:06:13] Runbin: There you go. Perfect.

[00:06:14] David: Yeah. I've wor Yeah. 

[00:06:16] Runbin: Yeah.

imagine like being able to wear like a couple order rings, and at least like being able to do telemetry on your hand movements and deploying that to like, I don't know, a million people. We'll be able to track how people move in real time.

[00:06:31] David: Well, that's what the Facebook band is, isn't it?

[00:06:35] Runbin: Is 

[00:06:35] David: They're a new band they've got, so when they, they're, they sold the, the display glasses. But they have the new band on your arm that they can now track your hand movement. Even like behind your back, you can do handwriting and it knows you're what you're writing. So it's really, really specific. 

[00:06:53] Runbin: Imagine now being able to track your hands in how you're fold laundry, how you're cooking, how you're doing all of these things with your hands, and that just becomes David's data set for how David moves, right? And now multiply that and compounding over like millions of people. That's a really good training data dataset for like humanoid robot companies like that will be a viable business,

[00:07:18] David: Well there is,

[00:07:19] Runbin: datasets.

[00:07:20] David: there are those people now who are building those robots and they're training and they have to wear these little silly hands

that are mimic the robot's hands. To, to do that. But at what point, and I'm, we're going off on a tangent, which I love, at what point is that a privacy thing? Like I don't know if I want meta of all companies to be able to track me like that.

That's a lot. That's very invasive.

[00:07:46] Runbin: So here's the, the counter to that. What if we say it's not David, or it could be anyone, it's de-identified. 'cause I'm not interested whether it's David or not. Right.

I'm much more interested in that it is a human being doing this action. And so what if I can start paying people? If you have this adapted to your hands and you can earn, I don't know, like a really good annualized salary, just being tracked to do these tasks. It sounds very dystopian. I know, but

[00:08:15] David: It's a little bit, but I, I could see 

[00:08:17] Runbin: robots to do things.

[00:08:19] David: Well, it it, I saw a thing on LinkedIn the other day. It was, it was like 2024. You were a prompt engineer. 2025. You were a vibe coder 2026. You are an orchestration manager. You 2027, you're unemployed. I was like, it's not 'cause we're training. We're, we're basically saying here's how to do my job. That's a bit cynical, but if you're doing heavy labor or whatever, you're literally training how, and like Mer, what's that company? Meco, I'd never say it right, but they're one of those AI training companies, and I'm a doctor and I'm, and this is the, the fastest growing company in the world. I think they went to 400 million in 18 months, which is nuts. But there, I'm a doctor, a radiologist, anything, and I'm training AI how to do my job. I mean, what do you think is happening seriously? Like, I know you're getting money now, but what is, what do you think you're doing? You're training the bot to take over, right? Does that scare you at all? You cool with that?

[00:09:22] Runbin: I think it's inevitable because one, the technology is moving that direction. And two, so here's a fundamental question, right? If you're not the one doing it, would someone else do it?

[00:09:33] David: Oh, sure. Yeah.

[00:09:34] Runbin: And so, so then I think the, the bigger question to ask here is what happens then? Like if we assume that this is an inevitable event. And what I'm realizing a lot more, because we use a lot of AI native tools at our company, we're

14 people, five full time only, and we're able to do just so much more because we have agents, we have, you know, assistants, we have, I have a AI chief of staff that does everything

[00:09:59] David: Oh, now I have more questions. I'm ready.

[00:10:01] Runbin: Oh, yeah.

So, you know, being, imagine, right?

Being in a sales call and then right after the sales call ends, you land, you land on a beautifully designed, lovable proposal that is in the form of an interactive website, right? Like for that particular opportunity. Like that's a real thing. And so like when, when these things are happening and the systems are in place to actually do it well at scale, then the, the scarcity comes into human judgment.

Yeah, so whoever's running that process, I have to trust that person, not an agent. The person has really good judgment. That person has a really good taste on how that proposal should look. Design wise, that person should know that the content actually converts, and so that judgment becomes even more critical.

So then this is getting on a tangent too, but if we don't give new talent, the newly grads. The real world experience to build up those judgements. Are we gonna see a skills gap very quickly? And we're already seeing

[00:11:06] David: A hundred percent.

I'm a big believer. If you don't have juniors, you don't have seniors. And I think there's, there's a lot of companies right now in every industry, but especially mine, that are very shortsighted in that I'm gonna save a buck now and absolutely create a massive gap, that skills gap in five years. And, and, and I'm taking the gamble, I guess, that in five years I get rid of my seniors too. And so, I mean, that's one of the things that sticks with me. Owner of Big Pixel software development. What does Big Pixel look like in five years? I would've said a year ago I was more nervous than I am now because I think you, and I would love if you, if you agree or disagree. I think I am learning as we've jumped deep into AI ourselves, that the fundamental humanness of knowing what to craft 

[00:12:05] Runbin: Okay. 

[00:12:06] David: Inherently human, your audience, whatever you're crafting, is still a human. And that human wants to be surprised and delighted. And AI is not equipped to do that. And I think it's at that point, that's that human need that of that, to surprise and delight is always going to be there. Now the how, the, the, the implementation, I think is gonna be robots. However you wanna define that. But I think at the core, and the other thing is when you talk about, oh, well in, in five years, you don't need software developers. But here's the thing, the people who hire us don't want to go to a bot and have them h coded.

Either they want to go to a human and have us use the bots for them. Do you agree or disagree?

[00:12:50] Runbin: I agree. And so my engineering team, my CTO and, and everyone who works on our products they, one the AI's not really there yet, whereby they can actually write production ready code.

That's the big gap. However it's available to write a lot of simpler codes. Right? And so they're sort of performing in the, in the capacity of directing that work.

So they

go to sleep and the coding agent just codes, and then there's a. Bigger lift around code review and MERG merging the code.

And so again, like the judgment layer comes in. And so what I see, and, and I would love for scale social to actually do this, is what if I can pay someone twice, three times, four times the annual salary because they have great judgment. That's not impossible anymore, right? And so what if I can start hiring great super contributors who can give outsized returns because the tools are great. That then fundamentally changes the labor market, right?

[00:13:51] David: For 

[00:13:51] Runbin: actually contributes to the issues 

and 

[00:13:55] David: Yeah. 'cause you have super producers and then you have everyone else. I mean, the super producer by very name is maybe 10%. If you're being generous of the workforce, that means 90% of the workforce just has to find another job, which, wow, that's scary.

[00:14:13] Runbin: Yeah, well if it's externality, so I run a nonprofit which matches students, college students to startups in the region, right? And so the reason we started that thing is because we realized that there's gonna be a huge skills gap. The question is who bears the externalities of. Because companies are not gonna do it, right?

You see all these layoffs and they're laying off mid-level people now.

[00:14:36] David: Yep.

[00:14:37] Runbin: And so when they do that, they're signaling an excess of labor. And if mid-levels are having difficult time finding jobs, what happens through the juniors? And so I think that externality needs to be solved somehow. In the us I don't know where it's gonna go. In other countries. I might have some kind of a bad.

[00:14:57] David: We are challenging ourselves as big pixel to, we are this summer we are going to get an intern. And this is our trial. And what we're trying to do is see if we can take someone's scrubby, fresh green, behind, you know, web, behind the ears green, all those fun, fun words, and turn them into an AI coder during an internship.

That's test number one, if that works, 'cause they're still in school. If that works. We're gonna, hi. Our next hire would be a super junior Dev Straight outta college. And we will apply those things we learned. The question is, is does it, we don't know if that'll work. I, I don't know. 'cause we only hire seniors because we move so fast

and we can't afford to train a junior.

But if now the training is, I just need you to know how to use ai. That's a very different train

[00:15:45] Runbin: Hmm. 

[00:15:46] David: Right. We don't need you to, the senior is in your computer. Right? I, he Assuming you, we can teach you how to use this Well, we should be able to produce good code on the other side. That's the hypothesis.

We'll see how it goes. Alright, so we jumped way ahead. This was awesome, but we need to back way up. Tell me about scale social.

[00:16:07] Runbin: Yeah, so scale social is a AI enabled UTC supply chain, so it sounds very jargony.

Imagine. 

[00:16:16] David: how that just rolls off the tongue right there.

[00:16:19] Runbin: Imagine going to a restaurant or a local bar. You, you see a QR code on the table and on the QR code it says, Hey, share a authentic gen, genuine moment that you've had here. And in exchange you get free drink free side. That's it for the consumer, right? And so effectively we are building the infrastructure.

Think of Stripe, how Stripe processes payment. We're building that for authentic content. And you'll hear the word authentic and authenticity like basically countless times throughout this podcast because it's a big deal. The reason is because we were initially building an AI powered tool that just generates stuff with ai, and this was last year. And very quickly we realized that, well, two things. One, we cannot fundamentally compete with the ones with way deeper wallets. Meta Google name,

your big tech giant. And so it's a, it's a, I like to think that there is a David versus Goliath moment, but in, in the face of capital, it's, it's kind of

[00:17:23] David: Infinite money is hard to beat.

[00:17:25] Runbin: And so we said, you know what? Let's not, let's not compete in that lane, right? And then two, we kind of telegraphed last year or the year before. If the marketplace is flooded with fake AI generated content, what's gonna stand out? So I was trained as a biomedical engineer. There's a core concept called signal to noise ratio. You want to maximize the signal and reduce the noise existing in every single system, right? So then the core question here is how do you increase the signal to noise ratio in what you do?

[00:17:58] David: Okay, so you're doing that now, what do you define as an authentic thing? I'm a guy who just wants a free drink and a free appetizer, so I get in there and I just stick my tongue out and I call that my authentic moment.

[00:18:13] Runbin: Mm-hmm. 

[00:18:14] David: Are you guys equipped to just, do you just throw that out or you just put it in a library and the restaurant manager throws it out?

How does that work?

[00:18:21] Runbin: Yeah. So we collect the content and we get the rights to that content,

[00:18:25] David: Mm-hmm.

[00:18:25] Runbin: right? And so that's very important because content rights is everything in this game. And then what we do is we put it in a server. At a point of capture. So imagine taking that video at that point in time. We use AI to evaluate your content.

It's a, if it's a photo, it's the photo, if it's a video frame by frame, and based on historical data on what content has performed well for that particular segment. If it's a restaurant we know what creatives work well. So we score it. Great. So, so not great. And so at that point of capture, I'm able to tell the consumer whether it's great or not.

And if not so great, then I can let the consumer know that they should try again. Right?

[00:19:15] David: Okay. 

[00:19:16] Runbin: So much of the noise comes after, like there's just so much content being created all the time. Which inherently introduces a ton of noise and everyone's trying to fix this in post editing. Same analogy with this podcast, right?

If I, if I just says, if I say terrible things right now, absolutely horrible things, then this will probably go into post. Our thesis is what if we can train and enable and help people capture really good content upfront?

[00:19:48] David: And so what does scale social look like? I call this my open, my, my open field question in five years, no major hurdles. What do you look like?

[00:20:02] Runbin: Yeah, we would love to power digital experiences whereby real consumers, these are customers. We're all customers of someone. Some brand where we say. I vouch and back this brand, like I stand behind them, right? And I stand behind them so strongly that I'm willing to give them content so that other people can stand behind them as well. And

we do this at a global scale.

[00:20:31] David: You almost sound like the anti AI slop company.

[00:20:35] Runbin: Yes.

[00:20:36] David: Is that by design? I mean, are you guys doing your marketing message specifically to sound like that?

[00:20:42] Runbin: So it's by design. It's by learning what works, what doesn't work. It's, well, I mean, Atlanta today we'll be giving a keynote tomorrow on this very topic, right? And the learning came from my past experience. When I was at IBM Watson, we had opportunity to build AI systems as well. And this is like early days.

[00:21:03] David: Mm-hmm.

[00:21:04] Runbin: Everyone. When they hear ai, they think about efficiency, automation. They think about everything that they can do without involving the human. And I remember giving a keynote like many, many years back, we were in healthcare, so the stakes are even higher, where I took the position of human based system design and I was in telemedicine.

So imagine you have a pain point, you're looking to talk to someone. Every single other company tried to build a bot to triage you. We took the approach of building a solution for our nurses, so it sits behind the nurses. So efficiency's being realized, but as a patient, you're still talking to a human nurse. So now imagine that principle, that philosophy being applied to scale social as well.

[00:21:55] David: Hmm. So restaurants is just your beachhead or are you Okay? That's very cool. So where is scale social in terms of where are you? I know you guys just celebrated a million ar. Which is very exciting. Are you guys still on a runway or are you guys self-sufficient? Where are you guys? LA in that

world. 

[00:22:17] Runbin: So from bootstrapping the business, I bootstrapped it with $50,000 of my own money. My wife was our first investor, actually, I had to

convince her since the beginning. And that was last year. We've broken even about four times. By the time we raised that round we decided that we need to go faster.

And so we started hiring people, right? And so we're burning cash right now. We're not profitable anymore. We're on track to do $5 million this year, and so if

we consistently hit our revenue targets, we should break even again in Q2, Q3 of this year.

[00:22:53] David: That's amazing. I love, that's a great story. And this is where I would usually do my, my free consulting bit. I think I'll take that offline, but you don't hear that story very often. That's what I think is so great. So you bootstrapped it. 'cause most of the time, and so I'm building Teela and we are not raising money. The reason I'm not is because I want to, if I do take money, I want it to be a, on the curve of, of some sort of hockey stick. Whatever hockey stick looks like. This is not a billion dollar idea, I don't think, but it's, it's rising. And I realize when I'm talking to all now that I've been sitting on this side of the chair table, whatever, that so many startups come to me when I, when I'm mentoring, when I'm talking, I know you do this more than I do, I think. You talk to all these general young people and they're raising money before they have any idea if there's a there, there. And like with Chile's case, there are many times when I'm like, is there anything here? And there's no downside if there's not. I've worked very hard, but I've had a blast doing it. So cool.

Right? But if I was in someone's pocket, 'cause I borrowed a hundred thousand from my friend's, family, an investor, whatever. Ooh, that's tough. And I, so I, I love your story and the fact that you invested in yourselves wonderful. Always a good idea, but you've been growing at a pace that you've been able to break even multiple times.

I mean, that's a very unusual story. That's very cool. That, that's a really long way of saying cool. But it's just, it's nice to hear. 

[00:24:26] Runbin: Yeah, I bet. As you build Teela, you can sleep so much better at night because you know your own skin is on the on, on the, on the

[00:24:33] David: It's just it. If it fails, it's just me, and I'm okay with that. I told my wife, I said to me, success is if I fail in a blaze of glory. It just is a bad idea. Cool. I've lived and learned. I am wildly successful. Yay. Very cool. It's the middle ground where you just kinda meh. You're just limping through. That's what I would be disappointed in.

Right? That

eh, we've got a few users, it's kind of here. That would be disappointed, right? Because then that's just like you didn't, I don't go big or go home, I guess is kind of, kind of how I'm looking at it.