Podcast Summary

This episode of the Cambridge Marketing Podcast features a discussion with Patrick Husting, a software entrepreneur and solo-entrepreneur. Husting discusses his use of AI technologies like ChatGPT and how he has leveraged them in his daily work and entrepreneurial ventures.

He describes using AI for research, writing, and programming, and how it has enabled him to be more productive as a one-person operation. Husting also shares the story of how he developed a popular horse management app using AI to aid his own research on horse health. Additionally, he provides insights into entrepreneurship, emphasizing the value of quickly testing ideas rather than overpreparing. Overall, the conversation highlights the practical applications of AI for knowledge workers and entrepreneurs.

 

Transcript

This transcript is auto-generated.

 

This week on the Cambridge Marketing Podcast, the Future of ai.

 

(00:26):
Hello and welcome This week we are in the exciting world of ai and if like me, you sort of know what it is, but you don't really understand it, my guest is definitely here to help. I'm delighted to welcome Patrick Husting, software entrepreneur and solo-entrepreneur. Patrick, welcome. We have to start by where you are because you've got the most glorious weather.

Patrick Husting (00:50):
Yes, I'm located just east of Seattle, Washington, so everybody knows Microsoft and Amazon. So I am probably 10 miles from the crow flying to my place between the Microsoft campus and probably 30 miles from the Amazon headquarters. But it's a beautiful snowy day today. It's actually kind of odd to have snow here, but it's beautiful.

Kiran Kapur (01:16):
It really does. Look, I have to say it's a grey and miserable day in Cambridge, but then I think I could say that most days to be honest. So can we talk a little bit about, it's always one of those. Tell me about ai. Where's a good place for us to start for somebody who perhaps doesn't really understand it, maybe still a bit scared of it because some of our listeners are still scared of the ai.

Patrick Husting (01:39):
Yeah, there's been dabbling in AI for quite a while and really kind of at the researcher level or in big tech companies and who could afford it, but it was very rough. And then in 2023, we all basically got introduced to chat GPT, and it basically allowed you to have a chat conversation with the ai. And what's powerful about it is we can type in natural language, actually we can type in with really bad grammar and somehow it will interpret and respond to us. And I think that's one of the powers of it, is its ability to, no matter what we type in, misspellings, bad grammar, it can give a pretty good response to it. So a couple years ago, I think it was an exciting toy at first for a lot of people. And now today, after a couple years, and especially with what I've been doing, I deal with a lot of people, a lot of customers around the world and just helping them get more out of the AI and fine tune their prompts. But it's really gone from a select few people really down into the masses of people. Now, would a person who's out farming or assembling a car or manufacturing something benefit from it? Not yet, but I think people that are in the knowledge worker business, we're generating content, we're doing research, we're doing marketing, we're creating content using content. It's a tremendous value to them and they shouldn't be afraid of it at all. It's not going to reach out and take over their lives or anything like that we see in the movies.

Kiran Kapur (03:31):
I think yes, I think I am still sadly old enough to remember computers coming into the office, the workplace, and I think honestly, people thought that a hand was going to come out of the screen and bash you if you got it wrong. It doesn't work like that. Presumably the chat just says, no, we'll try again. Or

Patrick Husting (03:50):
If you want to, we can definitely later on the conversation get into the whole Skynet thing or you remember the movie Terminator? We can get into that because I actually had this discussion on another podcast recently. I'd launched my tools two years ago, two years ago and four days essentially. And when I was on a podcast immediately here in the United States and they're asking me about it and I was just basically saying, we're just starting to use it. It's not going to take over our lives or our computers or send weapons a mass destruction ahead us, right? Two years after using it and what they're cooking up right now. Yeah, I do have a worried side. I really do. We can talk about that later,

Kiran Kapur (04:38):
Maybe we'll come back to that. I'm trying to not spare people about ai. So I think one of the biggest problems that I have and that the listeners tell me they have is just understanding. You say to someone, AI can do anything, people just dunno where to start. So how do you use AI on a daily

Patrick Husting (04:57):
Basis? I've been in the software industry over 30 years and I'm one of those rare people in software where I can come up with an idea, I can actually code it and market it, sell it. I have multiple skills. Anyway, I sold the company. And so I was at home and I basically had a two year, I couldn't go work in the space that I was originally. And so I'd been playing around with what they called at that time, GPT from OpenAI, one of his research project. And right when Chap GPT was coming out, I got this Clydesdale, Clydesdales are predisposed to a genetic disease in those legs.

Kiran Kapur (05:39):
So Clydesdale is a horse I should explain.

Patrick Husting (05:43):
And they're basically born out of Scotland anyway, so those lakes can have a disease associated to 'em, and I could never really understand it because the technical term is really long and I really can't pronounce it, but it's CPL essentially, and I could never understand the medical literature on it. So one day I asked Chad GPT if it knew what it was and it did, and then it comes back with this stuff that I really didn't understand. So then I asked and I said, well, can you refine that and explain it to me like I'm 15 years old? And it did. And when I read it, I went, oh, I understand this now. So it's like dealing with professors, right? They're talking at a certain level that nobody will understand because of their terminology that you're using, but you could ask the AI to basically take it down to a 15th grade level and explain them to you.

(06:37):
And it did, and I was shocked by that. Then I just started using, I was actually using it for research purposes for my horses because I got 'em probably five years prior to that, and I want to be the best caretaker of those animals. It can get expensive, don't take care of 'em. So I was using chat GPT essentially to research health related topics on that, and I was cutting and pasting into Word. And then that's where I ended up writing the tool that exists in Microsoft office today called Ghostwriter or Autopilot. And that's how I got started to use it. So for me, I was doing research and I could go out and Google it, but then Google, you get all the links back and you got to go through each link and then you got go between all the ads. It was such a pain in the butt.

(07:30):
The AI really allowed me to go deeper in my research. Now, the research tools that are available today, like OpenAI and Google, they both have a product now called Deep Research. So if you're doing any kind of research topics, those tools are just mind blowing to help you go deep on a subject, but deep in a subject and make it understandable. And I think that's where a lot of us struggle. We're always trying to learn, we're always trying to educate ourselves, but our knowledge experts sometimes get a little weighty with the terminology. Not to say that they're not trying to impress us, but it makes it hard to understand. So I think AI gives us the ability to bridge that educational level and allow us to learn more, I think faster. I think it's a really a powerful tool for everybody today that's trying to learn.

(08:27):
I believe we've had decades of using the term, well, let's just search for it. I think now because of where the AI tools are, well, how do you use search today? We use search to research something, right? We're always research something when we use search. So now we're going to do research first. So now you're going to go to your favourite AI tool of the week, and you're going to go there and begin your research there for your answers then to try to browse 25 links and websites, blah, blah, blah. The nice thing is the new research experience gives you access to those article links if you want to go to that.

Kiran Kapur (09:14):
So on a day-to-day basis, and we will come back to your horses because I know you produced a horse app, which I'm fascinated by as well. On a day-to-day basis, which AI do you use? Are there any particular ais that you use or any particular things that you do with AI other than research?

Patrick Husting (09:32):
Yeah, so I have this autopilot add-in that I've created. It sits in Microsoft Office, and that's where I spend, I spend a lot of my time in Word because I'm always kind of in a information gathering learning mode. And so that's best for me in office because I'm taking these findings and I'm saving them off in a document or a document that I'm working on in my autopilot tool. I might be researching a subject or maybe I've written something and I want to improve that because my writing skills have never been that good in my, if you ask my wife, she'll just roll her eyes on my writing skills, but it's amazing that I can select a paragraph within my Word document and then I can go over to autopilot. I can basically say to it, can you improve this paragraph and the grammar? And it does.

(10:30):
I can actually tune it and I can say, make it sound friendlier or enthusiastic, and it'll rewrite that kind of in my voice too, because that's what I selected. And so it really kind of helps me in my day to day-to-day where I'm struggling with trying to get the idea started When I'm writing, we all experience this writer's block and it just knocks that wall down. The other part of my day is just research. It might be just general things that I'm working on. And then I tend to use a variety of tools like from open ai, and then there's Anthropic Law. I use them for programming. So I'm a programmer, and in the past I had a company of 54 developers in my company, and today it is just me and the ai. It's scary what it can do, so it can help me programme. And then if I want to create a pithy post on LinkedIn or social media, I can kind of write generally what I want and then I'll change my style and tone within my tool and ask it to write something for Twitter or write something for LinkedIn and it'll rewrite that post for me quickly. So it's really, again, if you're in any kind of content generation mode in your career, AI isn't necessarily taken away from you. What it's going to do is just going to make you a heck of a lot more productive.

Kiran Kapur (12:04):
Thank you. So I think we have to talk about your horse app because that's such a great marketing story as much than anything else. And you said that you started your research using ai. So how did that go from I'm researching my Clydesdale horse because it's got congenital leg problems to, I've launched an app which now has many thousands of users.

Patrick Husting (12:26):
Yeah, sure, sure. So I got these horses eight years ago, and I was kind of done in tech, if you will, and 30 plus years. But when I got the horses, I have all these new friends, there's the veterinarian, there's a trainer, there's all these people coming, you're buying stuff. And I wanted the app to basically track all that stuff, and I really couldn't find anything out there. So I created it. And as part of this, if you will, new career of mine, a horseman, I, I'm constantly caring for taking care. I compete on two of the horses and I'm constantly doing something with it. So that app that started out the first year is just a hobby. I needed something for myself. It's actually the number one app in the world for caring and managing horses. So there's Olympians that use it around the world.

(13:27):
There's a couple in the uk, Olympians that use it, Russia, Australia, New Zealand, United States. It's really kind of fascinating, but there's well over 50,000 horses in the app. And so that's kind of my life, which led to me doing healthcare research out for my horses, which kind of ventured me into the AI world. I am a dabbler in technology. I'm always early with ideas. I like to experiment, I like to research. I like to try new things. And so when I built the AI ad in for Microsoft Office, originally I was just doing it within the browser and I just got sick and tired of copy paste into a Word document and reformat, copy paste, reform, copy, paste, reformat, anybody who does education and marketing knows what a pain in the butt that is. So basically that made me decide, well, I'm just going to put this in Microsoft Office because Microsoft didn't have anything and they weren't going to have anything for a long time.

(14:32):
They never mentioned anything. So I released Ghost Writer in February of 2023, February 1st, 2023. And that night I went to bed and I thought maybe a couple people would find it right. And a blog post was done by one of the tech journalists that night. And then the next morning I woke up to hundreds and hundreds of orders pouring in, and I was like, what's going on here? And I had to generate a key manually for these people to unlock it. So as I'm trying to generate keys for these people, the orders were coming in as fast. I've never experienced anything like that in my life. And then it was about a month and a half later, Microsoft announced that they were going to have some product in about a year, and then my orders just went to the moon, just crazy. And so I automated it all and it kind of self did it.

(15:34):
The neat thing about my AI products, ghostwriter, and then there's also Autopilot is it's easy to use. It compares and features like Microsoft and the other ones, except it's a fraction of the price, they're charging $30 a month. I charge for autopilot just $50 one time because there's some math and usage within office you can see. So just recently, actually just a few weeks ago, I did the count over 1.6 million instals of the add-ins. It just blows my mind. I had one thing left in my career to check a box on, and that was a lot of people using one of my bad ideas, and this one just blew it away. It's very humbling the number of people around the world that use it.

Kiran Kapur (16:25):
So from a marketing perspective, you said that a journalist wrote a blog post about it on the day that you released it, but that didn't just happen. What did you do to get the journalist to do that?

Patrick Husting (16:35):
Locally, there is a tech news organisation called GeekWire, and so they're the tech news for the Pacific Northwest and very well known here in the States for sure. And so I knew the gentleman there. I did two podcasts with him before he covered the sale of my previous company here. He's almost like a brother from another mother. It's like we've known each other forever, but we've only met a few times. I actually sent him a note and I said, Hey, I'm going to release this ad into Microsoft Office. There's no way in chance Microsoft's going to approve it because it's going to compete against one of their products that they haven't announced yet, but this is what I'm going to do. And so he wrote an article on it. He says, if they publish it, I'll push publish. And so he had it all written, shared it with me, and I said, okay, well, we're just waiting.

(17:33):
And that night they published or Microsoft approved it, and then I send him the note on the approval. So then he pressed published two, and then his article went worldwide and it was almost in every country in the world, even countries you wouldn't think that are very computer literate are using it. So because of him, I got a lot of traction because I was early. And then I have a unique place where my add-in sits. I sit in the Microsoft office, add-in store. Nobody ever thinks about that as a marketplace for their solution. But that's been basically a gold mine for me. If I had to market outside of that marketplace into the general internet. As you know, there's so much noise out there. I don't call it noise, I call it smog. And to get above that layer of smog to get attention is nearly impossible today in my head.

(18:35):
I'd like everybody to believe I had this amazing strategy and plan when I implemented this, but actually I was very fortunate when I released it. I released, I was the first one to release something for Microsoft Office, and right when chat GPT launched, I put out my thing. Timing is everything. And because of that, I'm indexed everywhere in search because of it. So I get a lot of marketing people contact me, wanting to help me. I don't have that problem because it's already there just because of where it sits today in Microsoft Office. It's a good place. They got a billion users apparently. So it's a great place to sit.

Kiran Kapur (19:20):
Absolutely. That's fantastic. So Skynet and Terminator, do you really want to go there?

Patrick Husting (19:29):
So no, nobody should be afraid of the technologies. 99% of the people are going to use it to better what they're writing or to do research topics, or in my case, if they got a little bit of programming experience, it's almost like they have 10 programmers working for them. So I think what it's going to do is if you have an idea, it can create an app for you. Now you just need to be good at puzzles, putting the puzzle together. So it's going to create pieces for you. You can put it together and publish your own app. You can do that today. It's actually amazing. So all these people out there that have an inkling of an idea today, I think AI will help enable them to bring it to market when they never could before. And so it could be a software idea, it could be a content idea where they're selling knowledge documents or whatnot.

(20:31):
I've seen doctors using it. I've seen mental health professionals using it to better things. So again, 99% of the people are going to be the good people and use it for the purposes it's meant to be used for. But there's that 1%, right? And that 1% live within all the big tech companies, which is absolutely frightening because they control the technology. So two years ago when I was asked this question, no way in my mind could I see AI being broke, right? But just about two months ago, they took the smartest chess ai and they put it up against Open oh one or whatever, their smart version, and allowed them to play chess against chess against each other. How cool is that? Right? Except what they did was for them to collaborate on this game of chess, they gave them access to some of the underlying files of the chess programme to move pieces and to track what's going on and whatnot. But the command was the same to both of them. Win the game of chess, win the game of chess, right? Told both ais to do that. So the one that's focused just on chess, I wish I could remember the name. It was dominating the open AI one because that's what it's focused on. So what did the open AI one do? It started modifying the underlying files to win. And when I read that, I thought to myself, oh, that's absolutely frightening.

Kiran Kapur (22:08):
So

Patrick Husting (22:08):
If we give it access to your computer system, if we can give it access to the files on your computer, if you can give it access to its network connection, if you can allow it to go out and manipulate things like that, and you give it a vague prompt win at any cost, you didn't put any rails on it, then in theory it will do anything to win. And so that's scary about it. So in all these big companies, OpenAI and Microsoft and Google and Anthropic, and now in China, deep Seek, and the other ones, when they give it access to the core operating system, I would be worried because most of 'em fired their safety teams. And that's just so that they can go to market faster because it is from the big tech companies, it's a win at any cost. Today they're spending so much money on this stuff, hardware and people, and these models, that safety is the second thought. So there are worrying scenarios of it. Is that today? No, not today, but soon.

Kiran Kapur (23:26):
So as a user, what would you recommend that a user did to protect themselves?

Patrick Husting (23:34):
The way the AI systems work today, none of them have access to your underlying systems. But OpenAI and Anthropic, they have desktop clients for Mac and Windows now, but they're still in a, if you will, a little sandbox. They can't leave that little box, but Anthropic can watch your screen as you're working. So over on the one right hand side of the screen, you could have their chat running, monitoring your screen, and let's say you're working on an Excel document and you have some charts and stuff there, you can go over to the chat window and say, can you tell me what this chart means? And they can visually see that and then start to explain it. Yeah, so it's watching you. It's they're going to take it further, of course. But as a user, it's kind of like when we started this podcast, the browser was saying, it's asking for your microphone, it's asking for your video camera, and you have to give approval for that.

(24:37):
That will be upon the user of the application again, to give approval or not. I could see tremendous benefit, right? You're in your emails, how many emails do you have in your email client? Thousands, right? And so how do you find it? You use search and then it comes back with 25, 50, a hundred emails. But if you could tell the AI system, Hey, find these emails and give me a summary of those, and it does, that's actually kind of useful, right? Could it be used badly too? Oh yeah, totally. But there's tremendous benefits. So it's like everything you're going to weigh the pros and cons.

Kiran Kapur (25:15):
I love the big smile you have, as you say, cheerfully. Yeah, there could be problems.

Patrick Husting (25:22):
Well, it is fun. I love this technology. Every day I use it. It depends on whatever my work scenario is at that moment, but it shocked me on how beneficial it is. I don't see why everybody wouldn't be excited about it. You'll always see people go, oh, ai, ai, bad, bad, bad. But they're only reading something that they read in a negative article somewhere. Unless you use it, it saves me hundreds of hours. It saves me tens of thousands of dollars. It's a tremendous benefit, at least to me. But I'm a nerd.

Kiran Kapur (26:00):
So thank you for that. Can I ask, and it's something I do ask most of my guests, how did you get from where you started to where you are now? What was your career trajectory?

Patrick Husting (26:19):
I was literally the dumbest person in high school. In college, probably the bottom 5%. That was me and never amount to much. But when I was in college, that's where I always had a computer though, at home playing video games. But when I got into college and I took my first programming class and the four month class I had done in two weeks, I of knew that there was something there on the coding side of it. So I went technology. And then when I came out, I went into an IT organisation. I love tech. And when your parents are always telling you, graduate from college, get a job, work to be manager of the people and blah, blah, blah, blah, blah. And so I went through that too, right? That's what we're all told to do, right? Go to college, get a good job, stay there, raise in the ranks, you'll be the manager.

(27:12):
You'll do really well. That was a lie. So ultimately, it's really about exploring what you love and what you're good at. So I did that same path too. And I became the boss. I became the boss, and it felt like adult daycare to me, and I wasn't creating, I like to create. So then I went on my own and created my own company. And I would say the first two, three years was a lot of fun because it was just me and maybe five people. And then it jumped up to 50 plus. And here I am doing adult daycare again, and I was fortunate to sell it. And then here I am by myself and me and my horses and writing one of the most popular AI add-ins in the world. You just don't know.

(28:11):
I think I just have some curiosity. The computers and software to me is an adventure. It's my Indiana Jones moment that's really corny, but it's also the expression of my creative side, and we all have that. It doesn't matter what it's in. And so I think if we can find a figure out a way to express our creative side and actually make a living at it, that's a good thing. Unless it's a really bad idea, you don't want to do that. But in my 30 years of doing software, I've created over 30 commercial products. And more than half of those were complete failures. Cost money, cost time. But I learned a lesson, and the lesson I'd learned is not to do that, but it taught me how to do it better next time. And there's so many people, and especially for the youth today, the 20 and early 30 somethings year olds, they're really taught to prepare all this stuff ahead of time before they do something.

(29:19):
And I think you just do it. And my best example of that is, if you remember, just back to the Olympics recently, there was that one competition of shooting, target shooting, and there was that one guy, I think he was out of Turkey, he's probably my age, and he shows up for the weekend and wins or whatever, hand in pocket, no glasses, and there's nothing on his arm, just him and his pistol. And he wins, and people are shocked and awesome. And then when you watch the clips of all the other folks who were in their 20 and thirties, they all had special glasses on and contraptions on their hand, and they had body fitting suits. Why? I have no idea. But they were all just so overly prepared that they got crushed by the guy who just did it. And all the people that I've talked, I talked to a lot of people about really bad ideas all the time, and I always said, you're never going to have the complete answer or prepare for the complete answer. Just find what the minimal viable product is to launch it, to see if anybody's interested. And if nobody's interested, it's a bad idea. Move on to something else.

Kiran Kapur (30:28):
And you've only produced a minimum viable product, so you haven't wasted a huge amount of time and energy,

Patrick Husting (30:33):
And time and money and AI can help you. If it's a coding thing, it can help you code stuff or generate the content. It can help you with some of the marketing. It can help you build your website. It can help do so much for you. It can basically be four or five people for you in your startup. So now you can test ideas better, faster, cheaper, and then throw 'em away and go with something new without much loss of money. It's brilliant. I don't know why more people don't try.

Kiran Kapur (31:06):
Patrick Husting, software entrepreneur, solo-entrepreneur, creator of an app about horses, and also ghostwriter.ai. Thank you very much indeed for your time and for your insights into entrepreneurship as much as ai.

Patrick Husting (31:20):
Yeah, appreciate the time.