Navigating AI’s Impact on the Open Web, Freedom, and the Future of Technology

Dave Lockie:
Hello, you’re listening to Emerging Tech, part of the Open Web Conversations channel and Open Channel FM production. And I guess we’re kind of co-hosting today. I’m Dave Lockie, and with me I’ve got Robert Jacobi. Hi Robert, how are you?

Robert Jacobi:
Hey Dave, nice seeing you again. Long time. It’s great to be back on Open Channels for 2026.

Dave Lockie:
Yeah, it’s, uh, I can’t believe it’s February already. And as we were just joking before we started recording, feels like a lot’s happened this year already.

Robert Jacobi:
It’s January is like the longest month in the world. It’s, uh, because all the planning for everything else happens and you have to get it done in like 4 weeks.

Dave Lockie:
So yep, that’s 90% of, uh, 2026 done then. Yeah.

Robert Jacobi:
If only.

Dave Lockie:
And maybe this is a segue. So we’re just going to have a bit of a chat about the AI landscape today because I think there’s so much changing all the time. It’s very difficult to pin down a fixed agenda much in advance. So to segue, I would say I feel like my 2026 has been significantly expanded in terms of time just because I can get so much more done. And I feel like there are now multiple Daves doing multiple projects simultaneously. So I’m wondering whether we have actually compressed all of 2026 into one month already in terms of work.

Robert Jacobi:
Yeah, the problem is now the expectation is that there will always be multiple Daves every single month of the year. That’s the problem. We joke and talk about how Internet time exists now. We have AI time, which is another magnitude of difference in what needs to get done and what needs to get accomplished and what the expectations are of consumers, businesses, et cetera.

Dave Lockie:
Yeah. Yeah. It’s, it is in equal parts exhilarating and daunting. That’s the tension I live with daily. How about you?

Robert Jacobi:
It’s the ever-changing environment is what, uh, A, is fun to get up for in the morning. Uh, this is— technology is the space that you and I are in, and theoretically everyone else who’s listening to this podcast is also in to some degree. And it, it— if they’re not, they’re.

Dave Lockie:
Probably in the wrong place because this is going to be a tech-heavy conversation, right?

Robert Jacobi:
That’s right. Uh, but, uh, anything that continues to engage the brain and causes just a tint of stress makes you work harder and keeps you going longer as an actual human being versus an agentic bot. And we’ll get to that, I’m sure.

Dave Lockie:
I’m sure we will as well. Yeah, it’s a lot. I think we can just reflect as humans that it’s a lot, but I often hear and see people talking about technology in this, with this weird reference frame of, oh, it’s changing. Things are changing. Do we want this change or not? But that’s the very nature of technology. If you work in tech, the thing that you must expect is change and honestly change at the fastest pace. I don’t know if you’ve seen that graph where you have the rate of change of technology, which is the fastest change, and then the rate of change of society and people’s behavior. And then at the bottom, you have the legislation and the policy and the regulation that tries to then sort of deal with all of that. And they move at very, very different speeds. And that, that gap is widening minute by minute at the moment.

Robert Jacobi:
This will probably get me in a little bit of trouble, but that bottom part on legislation is interesting because it moves differently depending on what region you’re in. And obviously, you know, and I’m going to compare to EU versus US legislation and how Legislation in the US actually, I think, is still like 1903, uh, regarding a lot of this technology. And at least EU legislation, for better or for worse, for right or wrong, is attempting to understand it, the technology spiral, uh, better.

Dave Lockie:
Yeah.

Robert Jacobi:
Not to say that they’ve gotten it right. Again, I’m not, not picking sides here. I’m just saying that it’s, there seems to be more action on the, uh, European side of the technology discussions than there is on the US side.

Dave Lockie:
I mean, you’re right. Thanks EU for cookie banners and for now KYC, meaning like every crypto transaction in the EU has to be KYC’d. I mean, yeah, I’m not sure I’m such a fan, but you’re right. They do seem to at least be doing something further. Although, you know, the US has got the stablecoin bill now as the Genius Act, and Trump seems to be more tech-oriented, or at least his administration has been, than Biden’s, who I suspect was struggling even in the early days.

Robert Jacobi:
Comprehension of the— yeah, totally agree. I just don’t know whether that means less regulation or more regulation or what that looks like on our end. And I don’t think I’d be the first person to say that, that we don’t know where it’s going to go.

Dave Lockie:
No, no, no, quite. Well, we’ve had the Trump meme coin, so that’s one of the worst things that could have happened.

Robert Jacobi:
Uh, you know, I’m, I’m gonna bash meme coins, period. Uh, they’re all the worst thing. I don’t care whose name is on it, honestly.

Dave Lockie:
Yeah, I mean, they’re the worst thing if you don’t understand the game, but like, the, the way that they behave is— anyway, we’re getting onto crypto. We should be talking about— we’re totally.

Robert Jacobi:
Going off track, but you know, again, this is supposed to be a banter and that’s where we’re going to wind up at.

Dave Lockie:
And, well, you know, it’s also all convergent, right? The expectation that one tech doesn’t affect another tech is something that I found like a really interesting component of the AI narrative arc. As you suddenly see, we’re going from, oh, we can write a funny poem to, oh, it’s driving Unreal Engine and creating like real explorable worlds on the fly from my photo library, you know, like this and that, you know, the same thing is applying to biology and drug discovery and protein synthesis. And suddenly, like yesterday, the stock market exploded because Claude released a legal, I can’t remember what it’s called now, but a legal tool. And so that’s where I think it starts to get very interesting, right? Just to speak back to those layers of change, like pace layer change. I don’t think it’s tenable that legislation technology continues to operate at that level when the AI can understand and navigate that Like, technology. it’s— that’s totally asymmetric because inevitably there are loopholes and workarounds. And I mean, that’s why your tax code is so big, right? Because it’s full of corporate loopholes and workarounds.

Robert Jacobi:
Yes, but AI will now find all those loopholes for you, right?

Dave Lockie:
And democratize them. Democratizing tax optimization or something.

Robert Jacobi:
Theoretically, until someone doesn’t like it. But I mean, your point is spot on that AI is moving all these things forward in places that us as super tech geeks are not necessarily thinking about. You mentioned pharmaceuticals, biology, taxes, legal legislation, all that. And those sort of, that, that long tail of AI is actually the majority of AI that most people wind up using, whether they realize it or not. you And, and, know, we can get into the whole open source e-commerce, WordPress, WooCommerce universe, and I’m sure we Um, but, you know, will. you know, hopefully we’re at least some level of the forefront of this but, discussion, you know, what are the ramifications down the road for everyone else? And how can we, in maybe even the smallest of ways, channel all that energy, no pun intended, given how many resources AI utilizes in certain ways?

Dave Lockie:
So what’s on your— I’ll give you two questions you can choose from. Either what’s on your mind most around AI? What is there that you’d like to banter about? Or like, what is your day-to-day AI usage look like? What things are you building or using? Or like, what’s the most kind of impactful change for at the you moment?

Sponsor Announcer:
Um.

Robert Jacobi:
I already forgot the second question, or first question, because I’ve already gone down the road on the second question. Day to day, because day to day, uh, AI is almost, almost, it’s, it’s getting there because I still have to actually type. I just can’t scream out loud into the ether, at least confidently, about what I want done. But a lot of it is taking away sort of that executive assistant role of my life. And I’m a very, very much first adopter, except in very close settings, because especially with tools like AI, the security, privacy, paranoia hits hard and I’m not willing to give Claude, ChatGPT, whoever, Gemini, blah, blah, blah, access to my calendars, to my contacts or any of that information. That’s not going to happen. But can I run discrete tests against content that I provided very specifically? Yes. And do I do that often? Yes. Do I want it to solve problems and create spreadsheets for me?, with discrete data that I’ve already vetted for, let’s call it, uh, Jacobi GDPR, um, purposes. Um, yeah, I’ll do that. Yes, data protection, I love that. Um, hold on, let me, uh, get one of my agents to get that domain name.

Dave Lockie:
Um, you don’t have them listening in and like actioning this automatically? Where are you?

Robert Jacobi:
Absolutely not. There are no devices in my household that are connected to the internet. I just don’t do that. Um, and I think that’s a good thing because it allows me to test an extreme case often. Like, why doesn’t this work? Oh, because the refrigerator is not connected to the internet. Uh, so that kind of AI does not need to exist in my life today, I say, because who knows what it’ll look like in 3, 5, 10 years. And, and maybe I’ll have my own little AI cube box that takes care of all this without any of those concerns. But at the moment, it really is— it’s a lot of those sort of like that grunt work that has to take place, uh, whether, you know, you’re in a sales role, marketing role, executive role, or whatnot. It’s just so much faster and easier to get basics done. And as a perfect example, just as sort of a hobby kind of thing, I need to redo my personal website, uh, and a lot of it is just like Hey, AI bot, go through the site, tell me the content that’s relevant, you know, tell me how to pull it out, put it into a, a, a CSV that I can import into a new site, call it a day. So that takes me 30 seconds to type out, ’cause I’m a slow typer. And then, uh, I have everything ready, you know, in 2 minutes. It’s ridiculous. And that kind of efficiency is, is awesome. And having been a developer for a long time, I can remember having to, I can even remember the SQL commands that I have to go through to, to get all that done. Exactly how I’d want it. And AI gets you 80, 90, 95, 99% of the way there. Fine, if I have to tweak a CSV for another 10 minutes, that’s still 1 hour of my life as opposed to a week of my life to get stuff done. Uh, and I— and everyone’s realizing across the industry, whether, you know, it’s, it’s in marketing, sales, development, whatnot.

Dave Lockie:
So let’s switch gears and talk about, um, your view on how it’s changing what we do for work. And I mean that specifically as in you and I, we work within the WordPress and WooCommerce ecosystem, but let’s also step back one concentric circle and look at, I guess, the open web as well. What are you seeing?

Robert Jacobi:
There are less humans on the open web than there ever have been. And it’s only getting worse. I mean, even where I’m at now at Blackwall, we do these internal reports on how much traffic we’re seeing, and it’s 55-60% of all traffic is bot agent driven at this point. So who is actually benefiting from all the content? How are you as an SMB, whatever platform you’re using, benefiting from the content? Are humans actually even interacting with you at all? And back, back in the day, uh, I’ll pretend I’m a lot older than you, that, you know, there’s this question, okay, we have to dumb down the internet, so, because people read it only in 30 seconds. Well, now you have like 0.3 seconds to actually read the internet because it’s not you reading the internet, it’s, you know, your agent, your, your RSS feed, your whatnot, that’s making decisions for you. And it’s small. The adoption around agentic is, is not huge today, but when everyone has an app on their phone that says, I’m your happy agent, let me do whatever you’d like me to do, uh, that traffic will all of a sudden take over 90% of the web. And, and how do we understand what that traffic management does for us you as, know, site owners, site developers, you know, e-commerce owners? Are we building out separate API-driven interfaces directly for those agents? Does that mean now we have to do twice the amount of work, but 90% of the work is only for the agents, and, but we have to spend way too much time on the humans that might accidentally visit our sites at this point? Those are the kind of questions that kind of bug me about, like, who’s actually using the internet? In a couple years.

Dave Lockie:
Yeah, I think that is a deep concern for sure. What came up for me while you’re talking is this idea of expansion and compression. I think certainly my experience at the moment is, sounds like yours as well, is the human input is often quite high level. I’ve got a high level idea or I have a high level intent. And maybe we can come back and talk about intent because that’s definitely my kind of primary mental model for where things are going. But we then kind of give it to this sort of like expansion system, which then goes and takes like, you know, an idea, like a couple of lines and produce a blog post out of it or like, you know, a graphic or something else. And then At the other end of the pipeline, you’ve got the consuming AI, which is trying to condense down all of this text or the image or whatever to what this person just needs to know to continue the task that they’re on. And I don’t think that’s always the case. Let me just roll onto this one and then we will stop and I’ll let you respond. But I wrote a post for the Automatic Design blog a little while ago. Around content for fun versus content for purpose. And I know I’m making balloons on the screen, but yeah, thank you, AI, making our day just a little bit better. And I guess the idea of that is sometimes you need a bit of information in order to continue your journey. And it’s really just like utilitarian. What’s the weather going to be like tomorrow? Or What’s the recipe for this thing? What you have on the other hand is beautifully written prose. Maybe it’s a book or it’s a piece of music or it’s a, I don’t know, a theater production or something that is worthy of consumption. The content, the consuming of the content is the end in itself. Sure, you might want a summary of a Netflix series just to decide whether you want to watch it or not. But when you sit down to watch it, it’s not a better experience if it’s just like a series of bullet points that has been compressed down for you. Whereas if you are deciding whether tomorrow’s a good day to go to the museum or go for a bike ride, you don’t sit there and watch the weather roll across the screen and sit there with your family and do that. And so That’s one of the things that I try and use when I’m thinking about how experience design is going to change, is that you probably do have those two different levels, and it’s not always clear. And we can see what happens when that gets perverted. So, like, a great example for me is looking for a recipe online, right? My intention is I just want to make, like, the world’s best chicken curry, for example. I search for it and then I have to read like 4 A4 sides, like a whole newspaper story about this granny’s old recipe that she found in the back of her wagon and blah, blah, blah, before you get to the actual recipe. Who’s reading that for entertainment? Nobody.

Robert Jacobi:
Nobody. Nobody.

Dave Lockie:
So anyway, I think that maybe that’s a helpful way to think about whether what you’re doing is something that you should get across as concisely as possible, or just, you know, it’s, it’s an experience you want people to linger on and not reduce down.

Robert Jacobi:
So, uh, I’m so glad you brought up cooking, and that I’m gonna assume that you actually— that’s a true story about cooking, because I’m the cook in the home, and, uh, that is my thing. And I honestly have plugins that when I go to a recipe site will just take me right to the recipe and not mess around with any of the fluff because I just need to know, do I have the ingredients? Can I do this? Uh, I, I, I’ve been cooking long enough to know that, oh yes, this makes sense. This is a good mix. And it’s gotten even worse because now when I just type it into Google, that summary is again, that 90, 95% good enough. Like, okay, so, uh, I literally just did this the other day. Um, I can’t remember what it was for. I think it was just for some silly roast chicken thing. But I had a question about whatever it was, and the summary told me everything. So I didn’t even click on a link. All I did was just go to Google, and Gemini gave me an answer that was like, oh, yeah, that’s right. I should do this low then high, not high then low, or whatever the weird matrix of information was that I needed at that time. But it was just instantaneously available. I didn’t have to like, you know, scroll through, scroll page through like an old-fashioned cookbook, of which I have way too many, or scroll, uh, through a bunch of stuff of granny stories. That, that’s great. And that’s great for a blog, but does, does that content become completely useless in this new internet? Uh, do we want Granny’s Cooking Stories? And if we do, would we rather just go to a site called Granny’s Cooking Stories versus I need a recipe?

Dave Lockie:
Hang on, Granny’s Cooking Stories. I’m seeing a Netflix series coming on here.

Robert Jacobi:
I’m seeing Automatic running a site called Granny’s Cooking Stories.

Dave Lockie:
And I guess for me, something that I’ve wrestled with, like my role at Automatic has been around the impact of technological evolution. So I joined to look at Web3. Now I’m working more on strategy more broadly, but of course strategy involves looking to the future, identifying trends, thinking how they’re going to impact their business, et cetera. And so I spend a lot of time thinking about the brand promise, I guess, of Automattic and of open source generally, which is around freedom. And thinking about how the context around that freedom has changed as the world around us has evolved. And I think freedom is a benefit, but you always have to weigh up costs and benefits. Like if something is freer, but 50 times more expensive or slower or harder to use, then like freedom isn’t enough on its own, right? Like freedom could be a differentiator if everything else is equal because, or like near equal, because then that’s like a personal preference thing. Like, oh, I prefer to use this, right? I’ll take a bit of a hit for it.

Robert Jacobi:
I can set up Claude at home, but that’s going to make me one of like whatever, 30 people that are going to do that versus everyone else is just going to go to, you know, any of the publicly available AI engines. I mean, that’s freedom, but it’s freedom with a price because, A, it takes time. It may take hardware that I may or may not have available. So that means I need to buy something. I may have a level of personal confidence and satisfaction doing that. Really, most people, 7-Elevens, they want to go to the gas station, put their car, whether it’s petrol or electricity or whatnot. They just want to be able to go click a button, get it done, because we have other things that we care about as human beings, uh, opposed to tech. You know, what does this open source technology stack mean? And that is the trickiest thing for folks to understand and why that matters. And a lot of people don’t think it matters, and they only realize when, you know, it’s, it’s like security. Security only matters when you realize that everyone else got hacked, but I didn’t. You know, that life insurance, it only matters until I die. It only matters until it really, really matters. And then when it really, really matters, it’s a catastrophe. It really is.

Dave Lockie:
Um, I don’t think we’re at a point yet where you can run something which is as close as it needs to be to the frontier models on like a bunch of Mac Minis at home. I mean, you can get like reasonably close, but I don’t think anything is quite at the level of Opus 4.5 or whatever it is. I think we will see the day where it is possible to run a local machine and to have something that’s as good as the frontier models are today. But then of course, the frontier models will have moved on as well.

Robert Jacobi:
And it’s the 80/20 rule. Is it good enough? If I buy my Apple TV in 3 years and it has a private secured LLM. Yeah, that’ll do the trick for most people, right? I mean, it’s sort of that, again, 80/20 rule about, you know, solves 80% of the— screw up the rule anyway— 80/20 rule. 80% of the people will be happy with, with 20% of the functionality. That’s what I’ll say.

Dave Lockie:
Something like that. Um, something like that. So I guess, um, you know, this is a big problem really, because the— for me, the— I guess the original sin of the the internet was not having two things baked in really. It’s really one, which is an open source equivalent to data, right? Because as soon as you needed a database, then you had to have infrastructure to run on it. It’s not like open source where people could just download it because a database that’s connected to a web application by necessity, it has to be a server, right? It can’t just be something that you run locally. And so as soon as you have a server, then at some point you need some hardware, and hardware implies like infrastructure. Infrastructure implies ownership and cooperation, and that’s like a very centralizing force, right? And we’ve seen that with the dominance of hosting companies within the WordPress scene, right? Like you can host tens, hundreds, thousands, tens of thousands, hundreds of thousands, millions of websites. The more you host, the better it is for you and kind of for your customers, unless you’re not incentivized.

Robert Jacobi:
Unless you’re totally slipshod, the fact that you’re gaining efficiencies by hosting more and more folks and understanding the issues with more and more folks.

Dave Lockie:
And scales of economy and like network intelligence and all of those things, right?

Robert Jacobi:
Correct. And that’s better for that next new customer because they will have already sort of acquired all that knowledge again by only spending a few dollars a month.

Robert Jacobi:
So Scott Galloway was talking about stuff, and I’m going to refrain from injecting any politics into this discussion because my life’s too short for these discussions.

Dave Lockie:
That’s a tough job if you’re invoking the Galloway.

Robert Jacobi:
Uh, but let’s say Prof G built a site, uh, regarding some political topic and, and lists a bunch of technology firms that you should be aware of for different reasons. For his sake, I’m going to just— I’m going to put it on Prof G’s, uh, shoulders at this point. Um, but then I, you know, being the smartass I am, I, I I do a, you know, what CMS on it, on that site. Well, it should be on WordPress. It’s not. It’s on a private SaaS company who may or may not actually be, uh, contrary to his aims in this discussion. And how do we get folks like Scott Galloway to be open source facing? Because they are very public in a lot of things and they’re missing that point on how open source and the tools they are personally using are not open source, which is, can be inferred as diametrically opposed to, uh, what he’s trying to say. Honestly, at this point, I’m thinking I should just reach out to Scott Galloway and say like, this is the conversation I just had on open channels and why are you not using WordPress?

Dave Lockie:
Yeah. I mean, I don’t know if he makes his own tech stack choices. I’m going to be honest. I think he’s probably more— See, that’s.

Robert Jacobi:
So this is where it drives me crazy. That’s the problem is like If I’m going to have a, you know, my car needs to have 4 tires, my technology should be open source. Those should be like the top level of, uh, requirements for anything I do. And I’m not saying I’m great about this because I live in my little Apple universe and I’m quite happy with my little Apple products, but I find them safer and more secure, and I have other reasons for utilizing them. And then you have a, you know, a, a punnett table about what makes sense and what doesn’t.

Dave Lockie:
Right. But, um, Cost-benefit right there. Like, you could be freedom-maxing with, you know, Linux, but you’re not because that would be a lot of sysadmin for somebody that just wants to, like, use the tech.

Robert Jacobi:
Well, and I only have so many resources in my heartbeat, uh, for whatever this lifetime is, and I need to also optimize those personally, right? And I think whether everyone realizes that or not, that’s, that’s part of the case., in everything we do is like, okay, do I really want to walk 4 miles for, you know, a sandwich that takes me then, you know, 2 hours to do, or do I call, uh, you know, one of the delivery services?

Dave Lockie:
Yeah. Yeah. Um, so where do you think the battleground or opportunities are for freedom? Online? If we’re saying that open source is like a valuable part of freedom, you know, it’s one, one component of freedom, right? Not the only one, but definitely one of them. Where is that fight? Is that fight happening? Is it even part of the conversation at the moment?

Robert Jacobi:
I’d like to think that a fight needs to occur between SaaS and open source. Uh, the GPL, as most folks are using, using it, taking advantage of it, does not force people to turn content over or, uh, works over back to the projects as long as they’re running internally. Uh, I will say Automattic obviously does a wonderful job of that because the work they do at.com and whatnot winds up, you know, significant portion of that winds up back in the actual WordPress project. So many companies are doing things internally or with SaaS that should be, uh, pushed back out, to the public as a whole so we can innovate and grow with them and, you know, and be free, to make technological decisions. And I’ll, I’ll even take this out of the WordPress realm. You know, we know who are good contributors or not. Uh, but you know, there are oil companies, there are industrial companies that are, you know, relying on open source tools. Uh, certainly the consumer product companies are, That stuff should be out there. If you’re using it internally and saying it’s SaaS and I’m following the GPL correctly, well, we need to change the GPL to, uh, enforce more accountability and allowing the public to reap some of the rewards that the public has put into those projects. Uh, because let’s take my family. It’s not just me who’s contributing, working open source. It’s the rest of the family who’s doing things. Their money and time is helping me do money and time to the open source project. So everyone in this household will benefit from that. And that only expands, uh, as you go out and to the rest of the world.

Dave Lockie:
Yeah, I have been thinking a lot about what an age of software superabundance looks like. Um, you know, we are I would say we’ve been in an era of optimization. So I’d characterize it kind of like this. Let’s use e-commerce as an example. So e-commerce was innovative to start with. Amazon was an innovation and PayPal and Stripe, these were all sort of things that emerged. And the whole idea of having a shop and selling online was innovation. And at a certain point, Innovation started to morph into optimization as people understood what a cart page was and should look like, and how it should function, and a checkout page, and a product catalog, and a product view page. And what are the patterns that people have become accustomed to and that align with the actual jobs to be done? So then we go into an innovation phase. And That’s where I think SaaS and proprietary can win, because in the innovation phase, what you really want is loads of developers hacking away and creating the patterns, and then emerges out of that the patterns that have the Lindy and can move into optimization. When you get to that phase, digital product practice is about being able to hypothesize, experiment, analyze, and iterate. It’s the classic kind of OODA loop, isn’t it? From like flying planes, like military and stuff, but just applied to software.

Robert Jacobi:
And then we start looking at shopping cart abandonment and metrics that were never envisioned necessarily in e-commerce 1.0.

Dave Lockie:
Yeah. And so my point about optimization is that SaaS and proprietary set you up better to do that because they centralize data. Like the more users and data that you have, the more effective the experimentation you can do and the faster you can roll those changes out, like the faster your OODA loop is, basically. And so that for me has been the last 10-ish years of e-commerce, right? It’s just been like optimization. We’ve seen Shopify and Squarespace and Wix like do really well. Because they can just take those patterns and optimize them. But AI, I think, brings us back to a new era of innovation because those patterns that were maximally useful and.

Robert Jacobi:
Well, just pattern, proprietary, yeah, theoretically.

Dave Lockie:
But I mean, like, the patterns themselves are changing, right? Because now you have, you know, we’ve gone from open web and Google search and ads to LLM conversations with no such kind of obvious place to stick ads, for example. So, you know, if people aren’t seeing ads and they’re not going to Google to do search, then how do you get traffic? And so the whole kind of thing is going to change. And we don’t know what that looks like, which means we’re in an innovation space, right? Because there is no established pattern for a lot of this stuff. And so I think That means that open source should win. Innovation happens better in freer and more open environments. I don’t think that’s about regression. I don’t think just going back to the— I don’t think we’re going back to the heydays of why Woo succeeded in the first place, for example, because it had X and Y and Z and was optimized for innovation for that era of e-commerce. I think it’s an era that rhymes but doesn’t repeat. So we have to think about what does meaningful freedom for both developers and customers look like in the AI innovation era. And for me, I think one of those things is probably, I think at the same time, that’s bad for proprietary and SaaS for at least a lot of them. And you could see that in the stock market now. Because you have, I don’t know whether you saw this Chamath, the Silicon Valley VC guy, he started this project called 8090.ai, which is essentially like, he calls it software factory. But the whole premise is if you’re an enterprise, you’re probably paying ridiculous money for SaaS per seat license fees, because that was actually the rational economic choice when building software was difficult and expensive, and you just needed to move fast and get things done. But we’re potentially, I would argue, we’re like now firmly in the era of software superabundance where those constraints and costs have changed. And so I think a lot of enterprises will just choose to build exactly what they need rather than paying for a prepackaged thing that does some of what they need, but is also subject to these constraints. So there, I think, you know, is that open source? Like, probably because you probably want to leverage the best-in-class open source components to build your overall system out. So you kind of drop down a level to like, what are these? Could be Python libraries, could be WordPress plugins, whatever it is. But you can like take the Lego blocks and then like build your own models instead of having to like— I’m.

Robert Jacobi:
Going to have to like totally interject it and just read because do developers care about that? Did they just run to— I mean, the amount of The amount of MVCs— but do the— well, LLMs don’t care unless there’s a, you know, what cost does it cost me, uh, more money to use, uh, you know, C# or Python. And, you know, that’ll be the determining factor in how those applications get built. You know, the amount of MVCs I’ve seen come my way that are, you know, something lovable.app, uh, is ridiculous. And there’s open source isn’t even part of the conversation. It’s what can I get done today?

Dave Lockie:
And the LLMs are using JavaScript frameworks, most of which are open source, like under the hood. Right. And that’s my point is you won’t necessarily, you don’t need to, we often won’t even know whether what is being built is open source or not, but LLMs will prefer to use all other things being equal, something that is like openly and freely available and licensable and to which they can access like the entire of the source code and change at will because it doesn’t involve of law and money and terms and conditions and like human level stuff. So we can see that.

Robert Jacobi:
So then what’s the, yeah, I totally hear you. What’s the incentive for humans to continue pushing content into that mechanism?

Dave Lockie:
How do you mean pushing content? Like you mean continue working on open source?

Robert Jacobi:
If it’s going to be eaten up by a machine and there’ll be no, I’ve been in open source for almost 30 years. So, you know, there are many types of rewards for open source contributors. You know, financial, uh, esteem, uh, you know, professional rankings, so on and so forth. If that just gets all eaten up by a machine and, and no, uh, awareness to those developers occurs across all the benefits that an open-source developer might want, why would anyone continue to participate in that process? That’s an insanely broad question, but I mean, I’ve seen the shrinking of open source contributors and communities over the last 3 decades.

Dave Lockie:
Here’s a good follow-up question. Do you think it’s important that people continue to be the ones that develop open source software?

Robert Jacobi:
Well, that’s the, that is the absolute logical next step. Will AI be the open source developer going forward? Optimizing. I mean, philosophically, humans talking to humans, it’s important for humans to be involved in the conversation. If we abdicate our, you know, creative responsibilities for open source to AI, then, then it doesn’t matter, right?

Dave Lockie:
Well, let’s talk about it a different way. You know, like, why does— why isn’t all software open source?

Robert Jacobi:
You know, at this point, I honestly have no idea. Um, it is So much of it is commoditized in the first place, um, given all the work you that, know, decades have put into making open source software. I you mean, know, thank goodness for Linux, thank goodness for PHP, thank goodness for, you know, C and yada yada. We can go on and on, JavaScript.

Dave Lockie:
You know, and we’ve, we’ve, we’ve even.

Robert Jacobi:
Seen clawbacks, the, uh, you know, Java being a wonderful example because that’s really where I got my feet wet in open source and how that got clawed back. And it’s just not part of the conversation as much anymore, uh, for most folks. Yes, uh, does a, you know, supremely optimized, uh, C-esque garbage collecting engine still matter for lots of the world? Yes, but, uh, so did Fortran, you know, 40 years ago, and that’s just not part of the conversation anymore.

Dave Lockie:
So whenever this kind of proprietary idea and pattern was introduced, like, why was it introduced?

Robert Jacobi:
Uh, because distribution— because distribution was constrained. If we go back to the ’80s and ’90s, pre-internet, you actually had to pay money for floppies, for CDs, for DVDs of content and all that. So there, there was an actual meaningful incremental cost to deploying software. There is no meaningful cost to deploy software on the internet right now. I mean, it would literally cost me— I could buy one cup of coffee today and to deploy a bit of software would take that for the whole year.

Dave Lockie:
So I guess the point I’m trying to drive towards is that there is a concept of proprietary software because at some point there was a cost to software, both creation and distribution. Like, you know, software developers were scarce and distribution was incrementally expensive. And so people started to charge, and if they were going to charge, then they needed to protect that IP in some way so they could make a living to continue to be able to like do the thing, right?

Robert Jacobi:
Like that’s the— and the buy side wasn’t there. Either. So if there are only— but these.

Dave Lockie:
Are human things, right? These are human things. Correct. And so I guess I go back to, is it Charlie Munger? Show me the incentive, I’ll show you the outcome. Where is the incentive for AI to do anything other than proprietary? I’d argue that proprietary is by and large like it’s an outcome driven by human incentives, not by agent incentives, not by AI incentives.

Robert Jacobi:
Uh, I’ll go, go resource utilization. Let’s think, think back to why music became almost free for a long time.

Dave Lockie:
Music is free. You can just You sing. can.

Robert Jacobi:
Just sing, but you know, and then we have the Napsters of the world distributing things and, would the cost for the companies that are running the agents be reduced if they could incentivize people to run agents locally, federated somehow, whatnot, those kinds of things. There, there are mechanics where AI might be like, hey, oh, the SETI project. You know, I ran that on my desktop for whatever, 10+ years.

Dave Lockie:
Did you find any aliens?

Robert Jacobi:
I, uh, no, I did not. I’m still looking. Even though the project’s been shut down, I’m still looking, you know, out the window.

Dave Lockie:
Um, maybe you did. It’s just they haven’t figured it out yet.

Robert Jacobi:
Uh, I guess my point being that there are ways we can incentivize people to get better product, and AI, because of that requirement, might actually lower costs and therefore increase profit. So I mean, there are, there are commercial mechanics that can be taken advantage of, uh, that accidentally or in this case, I’ll say purposefully engage open source and make open source a more meaningful part of the conversation.

Dave Lockie:
Yeah. And I guess on my mind is I think AI is bad for SaaS, for a lot of SaaS companies as we know them today. I think a lot of that gets replaced with like custom-built stuff that enterprises just build for themselves. That doesn’t necessarily mean that it is proprietary or open source because it probably doesn’t matter a lot of the time. Some of it’s going to be proprietary because it’s going to be sensitive, or at least the data that’s flowing through it is going to be sensitive. But I do think it does change the market conditions. And I think, yeah, as I said, we’ve kind of seen that in the market, right? Like if you don’t have defensibility as a SaaS, then, you know, it’s not a good time at the moment.

Robert Jacobi:
Maybe we just look at defensibility the wrong way. Being able to hide behind a legal shield is not defensible because laws can change very quickly and then all of a sudden, you know, companies disappear or whatnot. And then there are acquisitions and whatnot. You know, defensibility is better value for the, for the end user at the end of the day. And that end user can be internal or external, obviously, but I mean, the value being, you know, I— can I get it done faster, better, smarter? Have I improved on, on what I’m trying to produce? I like building things, and I think people like building things and getting things built for them.

Dave Lockie:
It’s a good time, it’s a good time for you and those people then, because we can get an awful lot built and done at the moment.

Robert Jacobi:
Yes, sir. And then there’s security and the hosting of it, but that’s what Automatic and Blackwall are for.

Dave Lockie:
Yeah. And I think that’s, you know, that’s another interesting thing to think about, isn’t it? Like if models, if intelligence is the new kind of thing, then like, you know, we go from tools to intelligence and intelligence systems, those also need hosting. Those models can be open source and open weights or proprietary.

Robert Jacobi:
Absolutely.

Dave Lockie:
Feels like to me that feels like the level at which freedom is most important at the moment. You said this yourself actually, just to call back to earlier in the conversation, you don’t have smart devices in your home. You don’t go load OpenClaw onto the root of your desktop computer because you don’t trust where that data is flowing compared to the value that you think you get back with it from it. But if you had your own bot in the corner and you were sure that that data wasn’t going anywhere else, then why not? Just connect your whole life to it because it’s acting in your interest, right?

Robert Jacobi:
Correct. And if I feel extra crazy about it, I can just take a hammer to it.

Dave Lockie:
Yeah, yeah, uh, might be expensive, but it would, would definitely do the, the job of helping you sleep at night.

Robert Jacobi:
Well, exactly. I mean, uh, I do think there’s always that, you know, security versus freedom argument, and, uh, I think we’re more free the more we control our security as close to us. I don’t need everyone else controlling my security. I’d just like to be able to control mine a little bit more, because right now we’re not incentivized to do that. We’re incentivized to actually give up all of our privacy and security to unknown entities.

Dave Lockie:
I’m struggling to remember which one it was, like the balance between it, but it’s often, was it security or freedom? But it’s actually more about convenience and free— it’s convenience and freedom that is the real thing that people choose against. Like if it’s just easier and it just gets done, then, you know, I’ll choose that.

Robert Jacobi:
Right.

Dave Lockie:
If the road to the prison is downhill and the road to nature is uphill, like a lot of people are going to walk downhill because it’s easier, right? Anyway, those people will go to jail and they might not like it. We’ll walk uphill together, Mr. Jacobi, into the sunny green hills of nature and freedom.

Robert Jacobi:
I just hope it’s not a Don Quixote moment. That’s all I worry about is my tilting at the wrong windmill.

Dave Lockie:
There is never a way to know except in the rearview mirror.

Robert Jacobi:
Thank you very much, Dave. On that cheery note, I hope everyone stayed for the whole hour because this.

Dave Lockie:
Yeah, thank you to the one listener that forgot to switch podcasts while they were doing something else.

Robert Jacobi:
Brilliant.

Dave Lockie:
Brilliant. Lovely to catch up with you. And yeah, that was for all you listeners, me, Dave Lockie, and Robert Jacoby having definitely a meandering AI banter about open source, freedom, and what it all means in the world of AI. So thank you for listening. Uh, this has been Open Channels FM.