Disambiguation Podcast Generative AI and Collaborative Content Creation - Transcript

Michael: Welcome to the disambiguation podcast, where each week we try to remove some of the confusion around AI and business automation by talking to experts across a broad spectrum of business use cases and the supporting technology. I'm your host, Michael Fauscette. If you're new to the show, we release a new episode every Friday as a podcast on all the major podcast channels on YouTube as a video.

And we also post a transcript on the Arion Research blog in case you want to stop by and read it. In our show today, we're going to take a look at collaboration and content creation using generative AI. I'm joined by Shawn Olds. Shawn's the CEO and co-founder of Boodle.ai. Shawn, welcome.

Shawn: Thank you, Michael.

Pleasure to be here. I appreciate you having me. Great.

Michael: I really appreciate you joining. Could you give us a brief introduction? Just talk a little bit about yourself and what you're doing with Boodle.

Shawn: Sure. Absolutely. So Shawn Olds, co-founder, co-CEO, started my career graduated from West Point as a computer science major and went into the army. And unfortunately, I couldn't jump out of airplanes properly, so the army kicked me out and that was the late nineties. And I got into startups and was very fortunate the first startup I did was not my own. And so, I was able to take a little bit more holistic approach to it and understand. Got the bug and pretty much for 20 plus years now have been building companies, both domestically and overseas.

I took a little bit of a break after September 11th. I went back into government service doing counterterrorism work, mostly in Southwest Asia and Africa. And then after that went back into building companies and was overseas actually in the Middle East, about to go to Singapore to do venture work there when my co-founder pitched me on the idea of Boodle, and I jokingly say that genesis of Boodle is my co-founder and I were lazy. In addition to everything we had done professionally, we had also spent 25 years serving on nonprofit boards. And so the genesis of Boodle was finding a way to bring technology, specifically data science and machine learning, to nonprofits in an easily consumable manner. And when I say easily consumable, we're on something similar to Zoom. Most of the people in the world have spent the past couple of years on Zoom or Meets, and I always tell people I'd be willing to take a bet with a good bottle of wine and a nice steak dinner that 95 percent of the people you've spoken to could not explain to you how video over IP works, but they don't have to because Zoom made it the press of a button. And that's what we wanted to do and what we did with machine learning and data science. And at the beginning of this year, 2023, with the advent of generative AI, we realized there was an opportunity to take everything we had done for the nonprofit to the sales and marketing community with predictive powered AI and add it with generative AI for a one plus one equals three opportunity. And so this year we launched Boodle Box, which is taking all of that, moving it forward for not just the nonprofit community or sales and marketing, but a much broader opportunity set.

Michael: Yeah, that's great. I've been playing around some with your products and having a lot of fun with it. It's an interesting approach and you certainly have a lot of different applications, which really could help a person, as they're trying to figure out how to do and what to do. It's not the easiest thing, I think. And we hear people talking about, we all need to be prompt engineers. And I'm just not sure I believe that. Maybe I don't know. So, one of the things that was interesting to me is was working through all of the different tools and content generation.

I think, I can see, I've talked to really interesting people about using the generative AI in marketing and product development. And there's all these interesting use cases. But one thing that I never really thought about until I started playing around with Boodle was wait, how does this work as a team? Like how I'm fine with my individual tools. But how do I work with other people? So, I that's one of the things that intrigues me about what you've done there. So, I'm really curious what you know, How? What inspired you to jump in on the collaboration side of it and content creation and talk a little bit about how that works in the product.

Shawn: Absolutely. So, what inspired us was just the natural evolution. As I mentioned, we realized everything we had done could be combined with generative AI, and it wasn't an epiphany on ourselves. We saw the interview that Reed Hoffman did with Sam Altman, where Sam opined that the next frontier beyond the large language models were going to be middle layer AI companies that have proprietary data and proprietary insights they could put on top of an LLM and supercharge a vertical or an industry. And we had built 35 billion proprietary insights on how all 250 million Americans consume and donate in this country. And so in very short order, by April, we had an alpha of that built and started to offer it. And while we were beating our chest that we were the first middle layer company in the nonprofit space and one of the first for sales and marketing, what we started to realize as we talked to our alpha testers was people really don't care where you are in the race. They really don't care how strong you are. What they cared about was, did I get what I needed out of my experience when I sat down? And as we scratched the surface a little bit more, some of the other things we started to find is people found generative AI to be very not easy to use.

It's very siloed, right? If you use ChatGPT and Bard and Midjourney and Stable Diffusion, You have four separate chrome windows open. The other thing we found is that the sharing and socializing of these prompts and what people are getting is not easy. If you read a Fortune or a Medium or a Forbes article and someone is sharing a prompt, you're looking at a screenshot that's been cut and pasted. And that's just not conducive. The other big thing, two big things we found, one is discovery was hard. The number of tech savvy, business savvy, name the savvy person you want that I have spoken to over the past six months who said, Oh yeah, I tried it, but I really just didn't see the value in it.

And the problem is you take busy people and if they try to spend five minutes putting eight Google searches in as prompts, they're going to get eight bad results. And after eight bad results, they're moving on to the next thing in their agenda and their calendar. My co-founder and I coming from a military environment, we've always worked in teams. We always we gravitate towards people where our strengths can overcome their weaknesses and their strengths can help accentuate our weaknesses and build something great. None of these AIs talk to each other. They're all amazingly powerful tools, and they all do it in silence.

And so that was what led us realizing that in order for us to optimize our own middle layer agent, we needed an ecosystem. It could operate in where people can share. They can collaborate. They can be able to get each of these AIs to also collaborate to give them the best answer that they need. And that's what got us down this road.

Michael: I got excited several years ago around the collaboration world and social sharing as it moved into the business world. And I could see the value and like you, I have a military background, so I was used to working in building teams and managing teams and I love the idea of the collaboration, but then moving into the world we live in today, I hadn't really thought about. And I guess part of that is because the use cases I had in the first use with the Gen AI tools was very isolated. It was I'm researching a thing or I'm trying to get an image to put up for a blog post or whatever, right?

But as I start to look at this in the context of businesses and talk to businesspeople, I more and more could see it. How this could be a team tool that really does need that interaction because, somebody is going to be really good at writing prompts and somebody is not, but somebody gets really good images and somebody else isn’t really good at that. And it's, so I can definitely see that. So, I'm curious from the feedback you've gotten from people using this, how are creatives using that today?

Shawn: So you saw in just this month, we added the first ever generative AI prompt feed social feed. And so part of that idea is helping that discovery happen faster, right? As someone comes in, they've never used generative AI before. They can start to see a feed of prompts of what people have been doing. And I look at it, I'm on there daily and even I will learn things. Like I'll see him like, oh, if you use that word in the image generator, it generates something entirely different.

And so where we see creators using it is just looking at what the community is doing. And being able to pull from that and say, Oh, that's how I want to incorporate this into my next prompt or the next thing I'm doing. We see this both at the education level. I've had talks with a number of educators who are incorporating generative AI into what they're doing, mandating that it gets incorporated. If you take a look at Ethan Mollick at the Wharton Business School or Dr. Kurt Beyer at the Haas Business School. They now make it mandatory to use generative AI in what they're doing so that their entrepreneurs are instead of trying to fumble through and come up with the 10 questions they're going to ask as they go talk to people in industry about a business idea.

Entrepreneurs for years have done this and it's about the eighth interview before you really refine your question set and get there. Now they can use generative AI to refine the question set. So every single one of the in person interviews they do is a valuable one. But now take it a step further, those 10 interviews, they take a long time to do. Imagine being able to take all the data and information from those interviews, feed it into generative AI, and now be able to interview the generative AI. All of a sudden, you've got 100 interviews that you can do. And so creators are using this as a way to get a kickstart. And to also expand what they're able to do in a more finite period of time.

Michael: I think of it a lot of times when I talk to different departments and functions that it is like a force multiplier. It helps you produce more, but I could see, in the context of being able to work with your whole team that is an exponentially larger force multiplier because of the fact that you can share across it. I've started using that a little bit this weekend. It's like a Slack-ish style social feed, which is really interesting. And people can follow and that sort of thing. How, when you see collaborative work and creatives working how are they using generative AI? Is it their handoff? Is it joint work around producing content, that sort of thing? Is that the way that this goes in the business application of it?

Shawn: So in the business application, I'm seeing three things happen. One is where people who may be better at prompts or good at prompts, or they just discover something because of a particular client they're working with they can now put it into a box and share it with all of their colleagues. And now their colleagues can propagate that takes a consulting firm. Or a venture capital firm, they can now take that and leverage that across a multitude of clients or portfolio companies. And it may be that something, someone runs into a use case they hadn't really thought about. Now they can put that into a collaborative box within Boodle and now someone else can be more proactive.

So rather than waiting for a company to have to need something, they can take something that's more proactive and help that company get to a place faster. A second way we're seeing it used is where people are able to come in and just fine tune things. So because as you saw, we've got a multitude of bots on there. Now someone can say what if I asked this bot to add some value? Yes, ChatGPT gave me this wonderful answer, but I know that sometimes Bard or I'm sorry, Claude can be a little bit more creative. I shared with you earlier that my daughter thinks Claude writes better bedtime stories, but you know that individual now can go take something that was a good product by one of their colleagues, but decide I know I've used this other bot or this other middle layer agent to help me do other things.

And they can add in, make it even greater and make it even stronger. And then the last way we just see it is as organizations are able to put boxes together, they can start to pass things on to new colleagues coming in. So imagine a consulting firm that is able to create its own data bot behind a trust and safety layer so they're not having to expose anything to an underlying LLM. And they're able now as their new Harvard MBA comes in that they're spending a lot of money in rather than that Harvard MBA spending a whole month going through meetings and having to talk to partners. Literally on day one can start interrogating their own internal system on projects they've done in the past year and be able to start adding value on day one rather than waiting several months to ramp up. This has been done before.

Michael: We've talked about this, with sales teams and with some of the startups that I work with about that same kind of thing in the context of sales you could use this to ramp salespeople faster because you could give them a tool that is almost like a tutor. It helps you understand the use cases of the product and the industry. And could learn from the successful execution of your sales team across the whole team. So that's, yeah, that's interesting.

Shawn: That was one of the first use cases we had with our middle layer agent. You hire that strong salesperson out of the Columbia MBA. And on day one, they walk into. Under Armour and they get the keys to their office, their parking pass and access to the Under Armour box, right? And in the morning they get their 401k set up, they get their direct deposit, all this stuff the HR person doesn't want to have to sit down and do, they now interact with Under Armour GPT to get that all done.

They go to lunch with their new sales team and in the afternoon they walk in and using Boodle GPT along with maybe Claude or some other middle layer agents, they're now able to say, okay, what was my worst selling skew? How many of our customers bought twice of that skew? What's the underlying profile of those people who bought twice? How many people in America look like that? What's the best way to come up with an advertisement on social media that's going to appeal to those people? And all of a sudden on day one, they've started a campaign to salvage one of their worst SKUs that's really powerful

Michael: And I, one of the surveys have said for the last several years, that takes about nine months to get a salesperson up to speed and productive. And that's a huge investment of time, especially in the context of the average tenure of a salesperson, at least in tech is like 18 months. So you really only have half of the time they're there that they're really productive if use those numbers. So if you could accelerate that's extreme value back to the customer into the business because you make the customer experience better. And then you also, make the revenue generator generate much sooner. That's impressive. Absolutely. So years ago, and I was just thinking of an analogy. So but years ago when I was at PeopleSoft, we used this old IBM tool. Actually, I guess it wasn't even owned by IBM at the time called Lotus Notes. But not for email. We did use it for email and it was, but it had this concept underneath that you could build out these knowledge based databases. And so from a consulting standpoint, we'd have you know, database that had a bunch of code snippets that you could use and one that had project plan samples that you could use. And it seems to me that's a very much an analog for what you've done with the idea of building the box. Is that accurate? Is it more? Could it be like a knowledge management tool? You think?

Shawn: It could, but we go a step further in one as it grows, the communities are going to become very big and it's going to be impossible to just sit down and searching or it's gonna be impossible just to filter through and go find them. So we've created a search capability where you can find prompts. You can find boxes or you can find people that are pertinent to what you want. Take it a step further, though, for that individual who's maybe not got a lot of experience or just doesn't have time to go through a thousand plugins that exist.

Now, as they start to type in their prompt, the system will recommend the various bots or middle layer agents that might be valuable to add. And we've created the marketplace where those exist, and they can decide to bring those in. And then going back to the collaboration point. When they save that prompt, and as we call a safe prompt poodle when they save that poodle, it shows all the eyes that went into it. So is their colleagues maybe want to recreate that down the road? They don't have to sit down and have a meeting to go discuss. How did you create this? They can just look at the Ais, add those AIs in themselves to a conversation and now engage and move forward on their own productive pattern.

Michael: So it really is a road map for that outcome that you're trying to get to. And you can see the whole picture, the whole process of it from the prompt all the way through the different steps that it took to get there. Correct. Yeah, so you're, so that is a much deeper knowledge transfer than you would see if you were just thinking of a knowledge management system in the past, right?

Correct. So the obviously I can see a lot of value here, but I'm just curious about the opposite side of that from a business perspective. What are some of the challenges that still have to be addressed to make generative a more accessible and more useful for creators?

Shawn: First and foremost is always going to be security. How are we making sure? And we believe that the trust and safety layer we're building and many other organizations are as well, is the first step towards that. Second is going to be responsibility. Using this information responsibly. One of the things we're discovering that is going to be invaluable to organizations are the prompts. We have told people in our alpha and our alpha users agree to it that, hey, we read your prompts because we want to make sure that we're building the right product. In the long term, we will disassociate the prompts with the individual. But those prompts still become very valuable. You think about it within an enterprise.

Think about how many Google searches a person does at a company every day during the course of their work. And those companies don't know anything about those. Imagine you're on the senior executive team and that a dozen different individuals, you don't know who they are, but a dozen different individuals have all asked for the same data. And they haven't been able to get an answer because of it. Your company is being held up and your teams are being held up. So now an executive can make a decision very quickly to get that data into the platform. So those teams can be productive. Flip over to the other side where a dozen people all make prompts about sexual harassment.

Now, there's somewhere in the company, there's a problem. And so now immediately you could decide on a Friday lunch and learn that you're going to hold a seminar. You have HR talk about sexual harassment and what people can do. And so you mitigate issues before they become big problems, both to the culture and potentially legally.

Michael: Interesting. That's definitely a use case I haven't really thought about is the idea that not only does it help you make people more productive, but it could also, from an executive team perspective, give you a different level of understanding about areas where there are issues or problems or things that you need to address in a much more proactive way than you would ever be able to with just looking in the rear view mirror, which is, of course, what most of us do these days to manage, right? It's very much a historical view. I'm trying to guess based on that versus I have something that lets me actually be predictive.

Shawn: That's really interesting, but all without sacrificing your team's privacy. Yeah, and that goes back to the responsible component of it.

Michael: Yeah. Yeah, that, that also is key, I think. And the idea of trust is really important when it comes to AI in general. And we've seen a lot of FUD in the world around using generative ai, using AI in general. And I think having some really clear, transparent way that says, look, we don't, we protect your personal identity. We're really using this to enhance the culture or, make the environment better for all the employees that, yeah, that makes a lot of sense. So let's put on our future hat for a minute, because this is exciting and you and I've actually talked about the fact that the compression and how things, fast things have changed are a little bit breathtaking right now.

What do you see for the future of collaboration and content creation. Where is this going for companies over the next, even two years is probably a lot at this point.

Shawn: You alluded to the first point a minute ago, which is ROI is going to become much bigger. We live, my father had two jobs. He was in the army and then he was a consultant. Those are the only two things he did over 50 year career. I say only they're very monumental things, but they were two career paths. We had our very first college graduate we hired. She stayed with us for three years, and when she left almost in tears.

She was like, I feel like I need to move on. I really don't want to choose, and if it makes you guys feel better, all 12 of the women that were in my sorority. Not one of them has stayed for three years, any place. Most of them have already had two, some of them three and one girl had four jobs in the span of three years. So people switch faster. And so companies from a, just a good point of view need to be able to figure out how do I capitalize on an individual while they're there? So I get that ROI. So shrinking that nine months that you talked about for a salesperson down to nine weeks, like all of a sudden, they're much more productive to the company during that time.

And frankly. They're learning more, right? Once an employee, whether they're a salesperson or customer success person or a programmer, once they get up to speed, they're learning more with each day and they're getting stronger. And so professionally, you're also helping your teams develop more futuristically as you look at it. I think it just makes people more excited to do their jobs, right? I if all of a sudden I'm accomplishing in three days, what used to take me three weeks to do, because I had to sit in meetings and I had to plan meetings and I had to wait for things. And all of a sudden I'm able to get things done faster.

I've got greater job satisfaction because now I can move on to the next thing and get things done quicker. Yeah. And then. Being able to bring about that collaboration even easier. We're in a society now that is much more seamless. If there's, lemonade to be made out of the lemons of the pandemic we've learned how to work more remotely. But with that remote, we need the tools that allow us to continue to work remote. And so being able to create these spaces where people can come and collaborate when they're not in an office and be productive in that collaboration is extremely powerful.

Michael: Yeah. That makes a lot of sense. And I know in that same context with the sales example that we did find in the survey work that in companies where they accelerate that onboarding and getting the salesperson productive, the less time it takes to get them productive, the more likely it is that they will stay well beyond that 18 month point. So to your point, if I'm if my level of job satisfaction That's is high than there is a much less likely that I'm going to be thinking about going somewhere else. If I'm learning and continue to learn, I don't necessarily need to go somewhere else to get more knowledge because I'm in the right environment.

Yeah. That makes a lot of sense, actually. I talk to a lot of businesses about using AI and one of the biggest questions I get is when should I get in, should I wait? because it's changing so fast. And so what advice do you have for businesses that are thinking about investing in generative AI and bringing that into their company?

Shawn: I think definitely, start exploring it immediately. And, whether you pick one tool, whether you pick multiple tools, the beautiful thing about generative AI, it's not deciding you're going to go buy a computer, for the person who had a small business 30 years ago, 40 years ago, buying computer was a big deal and a big investment. Once you got it, you're pretty much stuck with it for better or for worse. You were stuck with it. You can hop onto ChatGPT and spend $20 for the month and decide you don't like it and then go over to Bard and then go over it. As businesses, you could also have your team start exploring it for you.

All of these have free versions. Boodle has a free version. Boodle obviously brings them all together and allows them to collaborate and talk more. So in a very biased way, I'd recommend any business start exploring Boodle and using it. But I think the longer organizations wait. They're getting behind and their competitors are going to get more. They don't necessarily need to jump in, both feet first, and, go make a decision to integrate their entire platform onto one platform. But at least exploring it and playing with it and understanding more so from your leaders, what is going to be a good way to use it. And we see this in the academic sense.

I mentioned my co-founder and I both went to West Point had a good conversation recently with the dean of West Point of what do you, how are you looking at this? And his comment was, we're training the leaders of America. They have to understand how to use technology. Now, did he singularly have the answer how to do it? No. So he met with the department heads. You've got all PhDs that are teaching in each of the departments and said, we collectively need to find a way to leverage AI in a responsible manner that allows our future leaders of America to use it in a way where they still learn the fundamentals, but they allow their growth to become even more accentuated.

And so I think business leaders need to do the same thing. Rely on the people that they've hired. Have them look and explore how they use this in each of their departments and then figure out what the right solution is as they move forward.

Michael: And let's turn that around a little bit too from the actual employee perspective. And, obviously some of this is going to be driven by what their companies do, but on an individual basis, especially for creatives, what advice would you give them about how they could start to think about this and working this into their workflow?

Shawn: Yeah, I know a lot of creatives. Initially, their first reaction is I'm out of business. And one of the things we've always told people is the most singular, powerful AI team is the human machine team. The machine can do things that the human just could not do in certain periods of time. But at the end of the day, the machine does not have the experience that the human has. And so being able to take it fine tune it. And that's why that prompt engineering is so important. Being able to go in and Okay. Say, Oh, that prompt didn't work right, the generative AI did what it was asked, but that wasn't the right thing to ask. And so as a creative, what are the things that I can do to ask the right prompts that now get me catapulted that take, what was a week of preparatory work down to one day so that I can do what I'm good at?

I was at a conference recently and, ask people, how many of you have heard that AI is coming for your jobs? And, almost all the hands went out. And I told him, I said I hate to tell you're right. It is coming for your jobs, but it's coming for the boring parts. It's coming for the mundane parts. Everybody got hired into their jobs because they have a skillset. And that skill set is impeded every day by more ordinary mundane tasks we have to take care of. Now we can have AI take care of those mundane tasks and our companies now can get out of us what they hired us for and more of it.

Michael: Yeah I think that's great advice. I feel like that the idea of the enhancement that I get, the productivity gain that I get, the force multiplier that these tools can be is really exciting. And when you can see that in the context of your job and then also feel like you're contributing at a much higher level than you used to, because you are freed up to spend more time on the things that are more creative that do require more human interaction, that's just a benefit in my mind.

So that's. Yeah, that's great. That's all the time we have today. I could definitely keep this going for a while, but I think we probably don't want to lose the listeners at some point. So, Shawn, thanks for joining today. But before I let you go, I always like to ask is there someone that you'd recommend thought leader, an author, a mentor, somebody who really influenced your career that you think would be of interest to the audience?

Shawn: Since the topic was generative AI today, one of the gentlemen who's in, in the space most, and one of the guys who really gave us the epiphany of, I don't want to have 20 different Chrome windows open. I want to come to one place. I mentioned him earlier is Ethan Mollick out of Wharton business school. He, if you go on LinkedIn, you will get a steady diet every day of what generative AI is doing, what its weaknesses are, what its strengths are, how he's seeing the future business leaders that come through Wharton using it, how he's seeing other schools use it. He recently published a paper with about a dozen other PhDs. It was published under Harvard Business Review. But it did a study of using generative AI at Boston Consulting Group and the impact it had on its consultants. But if you're interested in generative AI. There's many people out there, but he is one of the ones that has been the most verbose in, in what he's doing. And I follow his feed on a daily basis to learn more. Oh,

Michael: great. Thanks. That's really good. And then actually I'm going to check him out. Cause I hadn't taken a look either. So the study sounds like something I'd definitely like to dig into. So like I said, it's all the time we have today. Thanks everybody for joining us. Remember to hit the subscribe button and for more on ai, we did a survey last month and published a research report that you can get on the area on research site. It's a free download, so free research, you can't beat that. So go check that out if you want.

And we have a, in the near, very near future, we're gonna have a comprehensive generative AI eBook that we're bringing out that really helps understand the use cases around AI. So I'm excited to hopefully get that out shortly. And next week you want to join us if you're interested in generative AI and project management, I'm doing another special edition from a Deltek Projectcon about all the things that they're doing and looking at to, to really help make a project manager's life easier, better, more productive.

That's it. I'm Michael Fauscette, and this is the disambiguation podcast.