The Bootstrapped Founder

Hey, it's Arvid, and this is the Bootstrap founder. So I saw something on LinkedIn the other day that stopped me mid scroll. It was a clearly AI written post. I could tell from the cadence, from the phrasing that was slightly off. It was super generic, inspirational, the tone of it and everything.

And underneath it were dozens of comments, super enthusiastic comments, supportive comments, and every single one of them was also AI generated. Bots responding to bots. It was horrible. A whole conversation happened there and not a single human being was involved at all, just machines. Every part of that exchange though was designed to look like it came from a real person.

And I find that quite tragic and hilarious at the same time. Tragic because deep inside of me is this wish for interactions to be genuine and authentic, things between people, exchange between real human beings. And hilarious because having lived through the development of these automated systems and having participated in using these systems myself to get what I needed, I kind of realized that we are responsible for this kind of stuff to exist. This is what people call the dead Internet theory, if you look at it from, like, the outside perspective. Right?

Just machines talking to machines. And I think it is worth talking about because we, as software founders and as entrepreneurs in this space, we're quite likely to use AI tools to build our businesses. The output of these tools now becomes often quite AI generated or has a lot of AI generated content or components, AI generated data, or AI augmented data, if you want to call it that. So in some way, at the very least we might be contributing to a network that is increasingly filled by AI systems talking to other AI systems with AI content. A quick word from our sponsor here paddle.com.

I use paddle as my merchant of record for all my software projects, and they take care of taxes, currencies, all that kind of stuff so I can focus on dealing with my users and my competitors. So if you would rather just build your product instead of dealing with all these other things, check out paddle.com as your payment provider and merchant of record. Now the dead Internet theory isn't about the useful kind of machine to machine communication that we have, like APIs or remote procedure calls if you live in the nineties, or systems that are supposed to interface with other systems just to fetch data, push data, create actions for us. All of that is fine. That's what machines are built to do.

So what we're seeing now is something different though. It's machines creating content that pretends to be human being engaged with by other machines that are also pretending to be human. And LinkedIn is the worst for this. Twitter is not far behind though. But the whole conversations there are happening where not a single authentic real human thought is present.

But everything is carefully crafted to look like one actually is. So let's talk about the quite obvious potential risks here. What happens when we as founders contribute to an Internet where only machines communicate with each other and humans are kind of left out of the equation? And maybe more importantly, how can we instead build tools that make it easier for actual human beings to connect with each other and still use things like AI? Right?

I don't want to say do not ever use AI tooling or generative AI in anything you do, but how how can we do this maybe more carefully and maybe more intentionally? The first thing that comes to my mind is this: if you know that your product can be used to generate slop which is any kind of content that spams forums or email inboxes screens and any content that only exists to elicit a commercial action but does nothing to facilitate human connection, well maybe you should rethink how much of this you put into your application or at the very least consider how much you try to prevent your users from misusing it right out of the box. Here's an example for this. Just take cold email generating tools. It's something that I've changed my mind on over the last couple months.

Like, I initially was completely against cold email, then I was completely for cold email because I saw it work using AI and all of that. And now I'm trying to find a middle ground here. And I think, generally, with all generative AI technology, that is the whole point. These tools are probably the target of a lot of scorn by people on the receiving end of cold emails, me included. I hate getting really crappy cold emails.

But even a tool like that can be useful, and even a cold email can be wobbly but still valuable and a valuable first step towards building a relationship with a potential customer at the same time. Question is, how does that happen? And does the cold email tool completely draft your email and then just send it hoping to catch a potential customer? Or is it just a first draft generator? Is it a tool that helps you write an email template and then fills in the specifics for each customer automatically but it requires you to review it and change it to put your personal spin on it before anything gets sent?

Is it really that important or even that smart looking at the bigger picture here to automate all of that away? Or could the tool take the first step while leaving steps for the human to complete? And I know this might be wishful thinking at this point. A lot of people buy these tools specifically because they automate everything away. They just solve the problem.

But maybe that's the bottom of the barrel that you don't want to be part of. If you have the capacity to create AI as a guiding system instead of AI as a pure generating system you're probably helping the internet not to become a worse place and you still use the technology that we have, like this magical tool, the LLM, the AI, to do something that your customers would otherwise spend a lot of time and money doing themselves in a less interesting version. And I think that's the core of the idea. Whenever you use generative AI, use it as not the end, but the means. It's kind of the categorical imperative.

There's a a lesser known version that Immanuel Kant used to describe this kind of relationship between why we do things. It's not just do unto others as you want them do unto you. That's the big one. But there is a version of this that says interact with any human as if they were the end as well as the means, never just the means to an end. And I think we should think about AI similarly here.

Don't use it as the output. Use it as the guardrail around the transformation from input to output. Facilitate transformation don't completely automate it. If you want to implement AI in your software business, in your tooling, find a way for it to be additive, not replacing. How can it multiply the thing that you and your customer is working on?

That should be the question. How can it amplify without taking away the act of creative work? It's a pretty relevant and substantial question and personally I found that in my use case for AI for all my products I always want it to be in the background. It should make the data that comes into the system that the users work within the system more usable by human beings, but it should not replace the work that human beings would do with that data. Case in point here is Podscan.

It's my business. Whenever Podscan pulls in a podcast episode, it tries to extract from the full transcript the main themes and the topics, the names of the people that are talking, the main people who are being talked about, and it makes all of the data available. But we don't do anything with it beyond that. We just extract it, which is where AI is best. Right?

It's finding meaning in data and then presenting it somewhere. But what to do with that meaning that is still on my customers AI is such a great tool at doing all this tedious non creative work so effectively so quickly like every week new models are being released that make it cheaper and faster to do this kind of stuff Unfortunately in a way these models are also better at being generative being creative and we use them for that because that's kind of cool but what we should be is we just use it for extraction analysis. I would love for this to be a bigger thing in our founder community not just the shiny generative stuff but the background stuff. The stuff where people want to just find data. They want to look at it and then think about how to use it.

They don't want their thoughts to be taken out. You don't want to externalize something that is a creative, meaningful process to you. You don't want to make a thing part of a process that's completely automated and abstracts away every single ounce of humanity. And if you offer this in your software as a service business, if this is the outcome of people paying you $20 a month, then you're actively contributing to that Internet theory becoming a reality. You're making things worse by making them easier.

It's really unfortunate, but that's something we need to grapple with. And there are other use cases for AI that are a little bit closer to where the user is. And one of my favorites in PodScan is something that is kind of smart, but it's kind of internal level work. It's the list similarity generation feature. Let's say someone has a list of like 20 podcasts that they've identified manually, 20 of their favorite shows or 20 shows that cover a particular topic that they found during search or whatever, they added it to a list.

Now the question is how can we find a 100 more 200 more something like that how can we make this list bigger without all the manual research because most of the important work figuring out the vector of the kind of list that we want has already been done And that's when AI becomes this background research agent, right? It can figure out what's similar between these shows that are already on the list. Do we have similar shows in our database? Are shows connected through the charts that they're on? What surrounds each of these podcasts?

What can we see around them? Can we kind of collate that into a list. And that's a lot of data acquisition, a lot of the value judgment happens on this broader, more general level, and this is what AI is really, really helpful for. In a way, that list expansion, it's kind of a generative feature. Right?

We're generating a list of similar shows, but that list doesn't get condensed into a newsletter or an article for a blog with, the top 100 shows or something. That's not what the product is supposed to do. People can export the list and do that themselves if they want. But what I want AI to do is to make the job of constructing this list easier so people have an easier time of then doing stuff with that data, not to turn that list into a finished piece of content. So if you can use AI to augment, to guard, to gates, to guide, these kind of things, I feel it will make things a little bit more human and a little less disastrous so that's my short exploration of the debt internet theory I could probably go on for weeks but I don't think we need to talk about much more than this because how we as founders could potentially be contributing to the dev Internet or taking that potential technology that we have that might contribute and funneling into a better perspective, that's what this is about and how we might step away from this make it less of a nuisance less of an annoying and authentic thing that's what I was trying to show.

The choice is yours. We can build tools that replace human thought and connection entirely adding to the noise of machines talking to machines maybe that's AGI, I don't know or we can build tools that give humans superpowers that really augment these capabilities while still keeping humans firmly in the driver's seat by still the human acting, the machine, the AI becomes a guide. I know which side I want to be on, I just told you, and I hope you will find this at least a little bit valuable as you think about where AI fits in what you're building. And that's it for today. Thank you for listening to The Bootstrap Founder.

You can find me on Twitter at avid kahl, a r b I d k a h l. And speaking of finding things, if you're a founder, PR expert, or part of a marketing team, you might be missing critical conversations about the brand, your brand, happening right now across millions of podcasts. It's out there. People are talking. And that's why I built PodScan to solve this.

It monitors over 4,000,000 podcasts in real time and then alerts you when anybody mentions you. You can turn all of this unstructured podcast chatter into competitive intelligence. And if you're searching for your next idea, check out ideas.podscan.fm, where I have an AI agent identify startup opportunities from hundreds of conversations a day so that you can then build what people are already asking for. If you know someone who could use that kind of insight, please share it with them. Thanks so much for listening.

Have a wonderful day and bye bye.