Bring Your Own AI: beinghuman.fyi and the Move to Thunderbird
Full transcript
Andrew: Okay. Brush the mothballs off from the equipment. Making sure it all works.
There we are. Of course, you realize you're now no longer being shot. I think this will be a big test. We're going to see.
There's definitely a height difference here. My eye line now is... Is your chair?
Is your chair? Yes. Did you know this is how you...
I know Sean's mini-me sitting here.
Sean: Did you know this is how you establish power, dynamic power, at a table is like, make sure.
Andrew: You know, I'm old enough. I don't need it. I'm okay to be the short one.
Sean: This does look silly. How do I sit here?
Andrew: Because I'm like, I can sit up.
Sean: Oh my God. So obviously we are not in our normal place. No, we have moved again.
We have moved yet again. And, but yeah, this is totally fun. So again, we've set up an impromptu studio because as it does, we continue to be in flux.
Yes. But also not just our studio. We as well.
Well, we're no longer in flux, but we were for a moment.
Andrew: We are. So we have, I can't remember, whether we've hinted at this in previous episodes, but we are now no longer, we're still at Arizona State University, but no longer in the College of Global Futures. And we are in the Thunderbird School of Global Management. That's right. Quite a major move.
Sean: We're officially Thunderbirds now.
Andrew: We are, yes. So, T-birds, Thunderbirds. And sort of squatting at the moment until we have the full studio set up done. That's right.
Sean: Well, this is also what happens. We're like, because we, you know, again, we've both been here for quite a while in terms of like coming through the building. Right.
I was teaching here quite a bit last year. And so there's all these little pockets of the building where I'm like, oh, we could do a podcast in there. Oh, we could do, what about that one?
And so literally
Andrew: And then you end up in this sort of bland conference room.
Sean: So literally we're just like, here's a podcast. We gotta set up somewhere today while we still work on getting our formal studio and stuff moved. Which will be a while.
Those take time. But the good thing is with a podcast, we can literally squat. And what this proves is we can squat anywhere and drop our gear and record something.
So here we go.
Andrew: So the big question will be whether we end up migrating back to my home office for a few weeks or whether we keep on trying to find Thunderbird places.
Sean: You know, we can keep on trying. Well, and we definitely will be recording something from the pub at some point because that is a crowning feature of the Thunderbird building in downtown Phoenix is the amazingly cool pub that sits on the top floor that has a beautiful overlook of downtown Phoenix and also just has a lot of cool character and stuff in it as well, which represents a lot of the Thunderbird history and stuff like that as well. Yeah, yeah.
Andrew: And actually, so as we get into this, and we'll say a little bit about Thunderbird today once we get into the show, but we ought to have a couple of people on in future weeks. Just say a little bit more, because this is quite an unusual school and institution. Yep, absolutely.
Sean: I've already been bugging the Dean, Charla Griffy-Brown, to come on. So she said, I will be there. So we will, you know, at a future episode, we will make sure we'll drag Charla on to give the proper sort of background and introduction, I think, to our listeners anyway, around Thunderbird.
Andrew: This is an exciting move. I mean, especially at this time with technology innovation and thinking about leadership and global management, because this truly is a global school. This is not US-centric.
Sean: Absolutely. And that's one of the things, I mean, I think there is, you know, all universities have their own feel, even within universities, different departments and schools have their own feel. And that's one thing I think for me that has always stood out is how special Thunderbird feels.
Because you walk through those doors and you do feel like you are in a purely, truly international. That's true. Institute of Higher Education.
A lot of big American universities that are very international still feel like a big American university that have international students. Yep. Which is in Thunderbird, one of the refreshing pieces too, is that you walk in, it doesn't feel like that.
It feels like you are in a truly international global mindsetted environment, which is a lot of fun. It is. It's represented in the students.
Again, the students that are in class are amazing and they bring such a diversity of perspectives and backgrounds and all that kind of stuff. You can see we drunk the Kool-Aid. That's right.
And this is going off the top of my head. I don't even have, I can do this like that. I don't even have any of the marketing material in front of me.
But it does, because instantly it does have that feel. And then again, also with the work with Future of Being Human, the work that we do, what an amazing place to be able to come and grow that work in a place that is already thinking about what are the tough challenges and how do we meet those tough challenges with leaders of tomorrow?
Andrew: That's exactly it. Yes. Yeah. So very exciting. And we'll say a little bit more,
Sean: but we should probably get to the intro
Andrew: dive in.
Sean: This is also the danger. We've got nothing where we're actually prepared. This is kind of like a quasi.
Don't tell them that. I mean, I mean, no, you're right. I've been, I spent hours.
Andrew: I can see you've got a couple of notes on your screen. Yeah,
Sean: these are old. These are what was
Andrew: on my screen.
Sean: Oh,
Andrew: now you're kidding me. Yeah.
Sean: Yeah. So, okay, we'll take a quick pause and then we'll just kind of jump back because we got a couple, I think a couple fun things to explore. We have. But also, hopefully this episode, this is, we've had a lot of heavy... All right, we'll be right back.
Voiceover: Initializing waveform.
Sean: Modulating signals for transmission.
Voiceover: Exploring the possible, probable, and preferable futures.
Sean: Welcome to Modem Futura. All right. I'm back.
Well, hey, a new background, but otherwise it's the same old Sean joined by his amazing co-host. Andrew Maynard. And that means you're listening to Modem Futura, the show that explores the intersection of technology, society, and the possible, probable, and preferable futures, asking deep questions like, what does it mean to be human in the future?
And again, I would say no better place now to ask that question than from the precipice of Thunderbird.
Andrew: School of Global Futures and thinking about what does it mean
Sean: to be a leader in the future?
Andrew: Absolutely.
Sean: And so, yeah, so I think a lot of excitement for us moving, you know, formally moving over. We've been doing lots, like I said, we've been doing lots of work with Thunderbird. I've been teaching for Thunderbird for quite some time.
But now it's formal. So we are here. They can't take it back.
Well, actually, no.
Andrew: After they listen to this, maybe they will. You have to find us in the building before you do it. But we should say just a little bit about what Thunderbird is.
Sure. And why we're here, because certainly from my perspective, this is a really interesting move. I started my academic career as a physicist and then did 1,001 different things.
And now I'm in what is effectively a business and management school, which is not usually where you end up as a research physicist. Yeah, yeah. But it also attests to the fact that Thunderbird is really very unique.
And one of the interesting things is, if you look at the last five, maybe 10 years of Thunderbird, there has been this very, very clear focus on leadership in a technologically advanced future and complex future. not only with digital transitions, but almost any other sort of technological transition. They were one of the first schools of their kind to embrace the World Economic Forum's Fourth Industrial Revolution framing.
And that is very much part of the ethos here. And now they're looking forward to what does it mean to be a global leader in an age of AI? So it jives very, very closely with the work that we've been doing over the last several years, both of us, around thinking about futures and foresight and leadership in a world transformed from what it used to be through technology.
Sean: Yeah, no, absolutely. And I think it's interesting because it was that Fourth Industrial Revolution framing is how I first actually learned of Thunderbird. Right.
Was through essentially find, again, being at Arizona State University, looking for, you know, who else is doing the things that I'm doing or looking at these things. And especially, you know, going back several years, that language was pretty scarce in terms of where you would find it represented, especially in higher education institutions. And so that was one of the things that I first sort of piqued my interest and attention to Thunderbird was seeing like, oh, there is an entire, wait a minute, there's a school that's dedicated to tackling those exact challenges.
So the things that were showing up in my keynotes, that were showing up about all these challenges, you know, that this Fourth Industrial Revolution era is bringing, you know, and I'm like, wait a second, there's a whole place that is kind of. Right, right. They weren't just sort of listening to your stuff and stealing.
No, it was like, thank goodness, you know. But yeah, so that was really, that was, of course, really interesting. And again, sort of that nice, that natural transition, because again, where does futures fit into all this work as well, I think is incredibly powerful because, you know, while we can say and leverage and discuss those challenges, the unanswered question is like, well, what is that mindset?
What is that skill set going to be needed to be able to see and handle that uncertainty that comes with this new era? And what better of a fit is that than futures? braiding that with the already existing amazing work that's coming out of the management element here.
So, yeah, I'm in leadership.
Andrew: So it's going to be interesting and exciting times.
Sean: Interesting and exciting for sure. Absolutely. But yeah, and again, you know, the courses are incredibly fun. I get to do a lot of the fun stuff.
Andrew: I've got to say, this is what is especially brilliant about this because you come in, you've got a full teaching load. So you're reveling in teaching these future global leaders with all your future stuff. I'm on sabbatical, so I just get to sit back and watch. That's true. There is no better transition to a new school than having the name, having the title, and just be able to sit there and do what you want for a year.
Sean: I think you've cracked the code. From now on, I'll be like, wait a second. Yes, a new transition.
Anytime there's a transition now, I'll be like, yes, but does that come with a sabbatical on the front end of it, please? But no, it is really fun. And I hope my students are catching on to this.
I mean, when I go into class, I'm just literally bubbling over. I think yesterday I had a class and my first thing I said is like, who's ready? It's a two and a half hour class or something.
And I was like, who's ready for the eight hours of content I've prepared? Because that's how much I am excited to try to, you know, we want to get there. What do you have, crickets?
Well, to their credit, they gave me a false cheer. And they were like, we are. And I was like, thank you for lying.
I appreciate that. And now we can move on and I'll keep it to, you know, reasonable content.
Andrew: But of course, that reminds me, One of the things that we're expecting here is a global audience for Modem Futura. That's right. Actually, maybe this is the secret, the cunning plan here.
This is how we break into the global market. That's right. I mean, we already have, I forget what, I mean, that was one thing.
Sean: We were already, actually, yes. It was 80, I want to say it's probably 90 countries by now, because it was like 88 countries like six months ago.
Andrew: But what we're going to do is we're going to expand from 90 countries with one listener per country to 90 countries with a few more.
Sean: Well, the real thing is like, you know, does the listenership bump up by the same number of enrolled students per class? Because, you know, that's the secret. That's the secret. It's now, you know, highly suggested listening for your syllabi.
Andrew: Of course, I assume that you have those assignments that say sort of the answer to this quiz is the fifth word in the third quarter.
Sean: That would be best. It's like what you have to do. As I have laid the hundred answers across 100 episodes.
You've got to find them. One word each. You have to listen to the whole catalog in order to ace this exam.
Yeah, that would be a fun trick, actually.
Andrew: Just drop that little answer in there.
Sean: Yeah, yeah.
Andrew: So if you're listening this week and you're in Sean's class, the answer is...
Sean: Digital global transition.
Andrew: There we are, yes.
Sean: Just bury things here and there, you know. That would be pretty funny, actually. I like this idea.
And it'll all be hidden under the guise of it becomes interactive. There we are. You know, it's a new pedagogy we're trying out.
It's interactive.
Andrew: Of course, the real question is, are they sufficiently up with AI to be able to use it to scrape the episodes and find it? That's right. Though I would say, actually, at the moment, there are enough steps you have to go through that I'd be pretty impressed if a student works out how to actually extract the text transcript using AI and then scrape it and find the answer. That shows a level of skill and dedication, which I think deserves some credit.
Sean: Absolutely. And that's because then you can start to split the group into half. You're like, who did the laborious way and had to listen to this?
And then the other people are like, yeah, it took me 20 minutes, top to bottom. I got it sorted because I used those tools. But yeah, and that's, but again, it's going to be a ton of fun.
And again, I don't want to talk too much because I don't want to do it a disservice to try to do my attempt at talking about Thunderbird and their history.
Andrew: And that's where it gets exciting, actually having some of the people here, the leaders and the practitioners to talk about what they do.
Sean: But just as a quick aside, so Thunderbird itself, while it has been part of Arizona State University for roughly 10 years, the school itself has been around for 80 because it predates, obviously, that merger.
Andrew: It's a really interesting path.
Sean: A really interesting path. And I pulled up their mission because I just wanted to make sure I had the years right. But essentially, it's been around for 80 years and originally chartered on April 8th, 1946, which was part of, it was built out in Glendale, Arizona, as part of Thunderbird Field, which was a World War II airbase that was used to train.
It has, again, this really unique history of training U.S. airmen during the end of World War II. But then also there was a lot of early Hollywood involvement of using the airfield to film a lot of these sort of like classic World War II era.
Andrew: But also wasn't there a lot of investment from Hollywood stars? Yeah. And this is where we'll have to get sort of one of the people that are really into the history of Thunderbird and talk about this.
Sean: Yeah, absolutely. And then also, I think it's interesting to talk about too, when we get to that, is also do a little deep dive. Because even the design of the building is a point that, again, as it relates to a topic we come back to often is about the care and intent and the design. And that's one thing you can feel when you're in the building, because the building is designed as such where each floor goes through a different geographic area as well as time throughout Thunderbird's history, which is really cool. So it starts at the very top where the pub is.
Andrew: Yes, but that's the earliest history. I'm sort of really quite bizarrely, you start at the top on the fifth floor, that's the earliest era. And each floor down you go, you get closer to the present. So when you walk in on the ground floor, that is the present and the future. The trouble is, as the future goes, they've got nowhere to go unless they build the basement down.
Sean: That was one of those things you were like, is there plans to keep, can we make the building taller? Like what happens? To jack it up, put an extra floor.
What happens in another 20 years when we need, like, oh, we need another four. Like, how do we do that? So I don't know.
But, you know, that's for future generations to decide, I guess. But, yeah. So anyway, yeah, super excited.
And I can't wait to have more people on to talk about just what that means. And, again, and for us to be able to, you know, as we do sort of unpack these things without pre-thought. And just the excitement that bubbles over of those new things, those new sort of things that come up that we just kind of gravitate towards.
I think so, yes. And have that excitement bubble. But the other thing I wanted to talk about, again, sort of in this mini-sode, is something else that we've kept teasing up from bits in here, little bits here, little bits there, was this, essentially, this AI slash web project that can fly under the banner of a URL known as beinghuman.fyi.
We have kind of slowly been mentioning this in conversation. And even in our Bill Gates AI manifesto memo episode, I actually put it in the show notes formally. So it is out there.
Andrew: Okay, good, good. But we should talk about it Because I haven't found any other organization doing anything quite the same.
Sean: No. And one thing I would say, so again, how listeners can engage with it, well, one, you can literally just go to, it is human readable. You can go to this website. Again, it's beinghuman.fyi.
And so you can see it. And at first you can see it and it interacts as a human reader. You can see it's a webpage.
You can look through it and you can click and you can read. But the thing that I think is that I have a lot of pride in this endeavor is that this, I think, marks something that is going to be something that we will see increasingly in demand, which is websites that are not only built to convey information to your human audience, but also AI first in the way that the corpus and the information that's behind the site is intentionally designed to surface the content, the intent, the ideas, the work, to feed specifically the hunger of these AI systems when you're asking at questions. So one of the things that I think is so brilliant about it is, one, we put our work in it, and we summarize and categorize and put this whole corpus, and you can talk at length a little bit about how that process was done. But one of the magic pieces about it is when it sits on the surface level, you can, of course, again, as a human, you can go to it and read it yourself, and then also use it to interrogate what does the future of being human do? Who are we? What's the
Andrew: point of all this kind of stuff? Why do you? What are we doing here? Why do you two fools exist?
Sean: What's the point of this? But then also behind that, though, is the incredible ability for it to feed in a way that makes sense for artificial intelligence to, again, whatever that might be, to be able to pick up the information from here and use it as a reliable source from the source.
Andrew: So you can literally go to almost any AI chatbots or platform these days and say, check out beinghuman.fyi, what the heck is this about? Or ask it anything. And it has the map when it goes to the website that'll allow it to pull out the information which is going to be relevant.
Yeah. So then, yeah, so there is a backstory here. And the backstory is one of the things I hate and detest is creating glossy brochures for initiatives that tell the world all the amazing things you've done, where you spend about $30,000 on sort of graphics and design and everything.
And it's all promotional stuff that as soon as anybody gets it, it goes straight into the trash. Nobody reads, I'm sure, those glossy brochures. I don't. As soon as I get them, they go straight into the trash.
And I was sitting there thinking, we've done amazing things, but there's no way on earth I am going to spend all of that money creating something that is just gloss that is going to go in the trash can. So what do you do that's different in an age of AI? And the next step there was thinking, well, actually, very few people read stuff these days.
They go to their AI and they have a conversation. So what we really need is a resource, which is AI legible. And as soon as you do that, you get away from this idea of just putting out sort of performance metrics.
And you can put everything from what you've done to your philosophy, to why you've done it, to what your vision is, to what your thinking is, to everything.
Sean: Yeah, to even why we don't do glossy magazines.
Andrew: So you can even go to your AI and say, have a look at beinghuman.fyi and tell me why don't they produce glossy brochures. And it will give you a really good response. So that was the idea.
And this was a coding exercise using Anthropic's Claude Fable. It was built actually over a couple of weeks in the end. And so this was not an overnight sort of thing.
But it was a case of sitting down with a coding environment and saying, this is what we want to do. How do we do this? What does it look like?
What is the structure? What are the documents? And over those two weeks with conversations between us and with us and Claude, basically a whole brain dump of everything that makes what we do what it is and what makes us what we are was put into this.
So now you've got this corpus of documents that are in this sort of hub and spoke sort of structure. So an AI can actually find its way through. If you were to print all this out, and this is the smart bit of this, there is a file.
If you really wanted to, you can download the whole corpus as a single file and feed it to something like NotebookLM or whatever Gemini are calling it these days.
Sean: Yeah, Gemini Book or whatever.
Andrew: It's incredibly versatile. But if you were to print that out, There are about 160 pages of stuff. This is a substantial piece of work.
Sean: Yeah, no, and I think this is incredible because I think, you know, two, what I love so much about this project also is I feel like, well, one, let me just say three things about it. One is the meta analysis of it, which is again, you know, to toot our own horn, this is another example of where we are pushing the boundaries of, you know, again, I think, you know, fast forward six months, 18 months, this will be the norm. As it is right now, it is not.
And so I do, from that sense, we are, you know, again, to sort of pushing the ourselves on the back a little bit. But again, but doing it in a way that is, here's how we did it. You can copy it.
You can do your own version and you can take it and run with it. But also it solves two interesting problems. One is, again, and I would encourage you, go ask it.
Why don't Sean and Andrew put glossy magazine, you know, reports out there? But also knowing that that in some ways that I don't want to say it hurts us. But oftentimes when you say like, who we are, what do we do?
And they're like, well, where's your glossy?
Andrew: Yes.
Sean: And you're like, well, we don't have that.
Andrew: Where's your PowerPoint slide deck?
Sean: Where's your pitch deck? And you're like, no, no, no, you don't understand. That's not what we do.
And then they're like, yeah, but why? And then there's always that awkward dance of having to explain the philosophy of why and where the value is. And again, this is a small operation, right?
And so we don't have teams of people who are just, you know, spinning that wheel for the sake of that's an expected output for a lot of, you know, a lot of, you know, think tanks and centers and initiatives that do that and nothing against them for doing that. But ultimately, most of it, you go, nice. And there it goes. So one, it solves that problem. Why don't we do that?
Well, you know what? Hey, you don't ask, you know, just ask. You don't have to ask us. But then two, also to meet that need of fewer and fewer people are spending the time to read and dig deep into the web content, and more and more is being diverted exactly to some kind of AI tool, whether that be an intentional application of you. That means you go to ChatGPT or your tool of choice and say, tell me about this. I want to learn more. What is the deal with those people?
Are they legitimate or is this something else? That's one way. Or even just also in terms of the web search now is driven by the Gemini AI overview. And so even when those, you know, what is gone is that sort of old school version of the web crawler. While it still will present those results, they are buried and pushed down because everything takes precedence in the AI sort of overview. And so how better to make sure that not only your organization or your information shows up in the way you want it to, but also making sure that you're giving that search
Andrew: engine, that AI overview even. The stuff that it needs. As you would say, access directly to the
Sean: metal, right to the source of the content that came directly from, in this case, from Future Being Human initiative. We put this information there. It is intended to be ingested both by humans and by AI first even. And that I think is something that is really cool that we have been able to do and do it in a way where you also, it doesn't lose the quality of just a traditional user interface for a website for humanized consumption.
Andrew: That's it. And just to emphasize that, if you go to the website, you can see you can actually access every single document in an easily readable form on the screen. And if you really want, you can actually download them and bring them off. Everything is fully accessible and incredibly
Sean: versatile. Yep. And also what I like about that method of displaying that information, having it available, is also you can fact check your AI response yourself. You can say, really? Did that guy really? No, come on. And then so you can go and look yourself and you can see that you can see exactly from the provenance of where that AI tool is grabbing the answer.
Andrew: I've got to tell you a story there with putting this together. So this was an interesting experience in itself, putting it together, using Claude Fable. So using a top end model for this, but it still kept getting itself in a rut that I had to pull it back away from. So one of those ruts was that initial conversation. I told it two things. I told it, we're not really interested in key performance indicators, KPIs, and we want to make sure that people actually will validate what they get. Don't just trust the source. And this was part of a broader vision that I had, but Fable latched on to both of those. So the first several iterations, you put it into an AI, and it will come back and say, well, I'm not sure about this. You want to check this, and you want to check that, and do not trust this site. I was thinking, that's not exactly what I was trying to say. So we had to sort of fine tune it against that. And then it got obsessed with KPIs. So it basically sort of said, Sean and Andrew are anti-KPIs, and this is why they will never touch on. I had to go back and say, no, no, No, we actually do count numbers where they're relevant.
We're interested in the numbers of listeners. We're interested in publications, readers, et cetera. That's right.
It's just that that doesn't drive our work. So that two weeks was a lot of fine tuning to make sure that it truly represented what we do rather than what the AI was trying to push on us.
Sean: And that struggle is also such a great representation of the real struggle that we feel. Right. Because it is, it's sort of like that struggle of, again, it's like the world is optimized for KPIs.
You're like, yeah, but that's not like we, those are important to understand where things are hitting and where they're not. But like, that's not the driving factor, right? But then it's like, yes, but that is the driving factor.
Andrew: You're like, no, no. And it's so revealing. There are so many things that I learned from this. But one is that virtually every AI platform at the moment uses templates.
So you ask it something and it matches it to a certain response template. And many of those templates, business templates, which naturally sort of veer towards things like performance indicators are important. Therefore, you've got to have them.
They're all very traditional, very conventional. So actually, if we're not careful, AIs are pulling us back into a very conventional world.
Sean: Yeah, oh, absolutely. And that's that fear of that regression to the mean again, right? Exactly, yes.
There's always that ever-present danger where if you are on the edge or the cutting edge or pushing the boundaries or pulling, you're pulling that, you're skewing that, you know, nice, rich bell curve, you know, there's going to be a tension. It's like, whoa, whoa, whoa, you're out of line, pal. You need to go back this way.
Andrew: So, okay, this is a good test of different platforms. And the other part of this is testing it out on all these different AIs. So I can tell you that the garden variety version of OpenAI is actually okay now.
We had some struggles there. One of the struggles is that a lot of these AI platforms, You give them a website. They won't actually get stuff back from it unless it's effectively in their database.
So initially, we had problems, including with ChatGPT, that you'd say, check beinghuman.fyi and tell me about this. And it'll come back and it'll give me loads of stuff. And I thought, that isn't right.
Then I'd ask it, did you actually check the website? And said, no, I went somewhere else. So we had to work around that.
It's actually working fine now. So ChatGPT is working okay. Claude is working well.
Haiku is great. Their lowest model. Opus is okay.
The middle one,
Sean: what's the middle one called? Oh, jeez. I don't even know.
Andrew: I never use it. The middle one is so patronizingly... Sonnet?
Is that right? So patronizingly supercilious, even with this, that I was finding myself wanting to literally throw it out of the window. Because he would sort of read the website and now he can read it.
But too often it basically sort of said, you know, this just feels like self-promotion. It was written by these people, so I really don't trust it.
Sean: Oh, yeah.
Andrew: So I'm not a great fan of Sonnet at the moment with this, but you can still get good results. You can. Gemini is bleh.
Sean: Well, I think that might be across the board.
Andrew: The one that I enjoyed using the most was Grok. Grok is just beautiful with this website. I'm sorry to say, because in some ways I'm not a fan of Grok, but it does work very, very well.
Sean: And I will say, I have never used it. I have never, that's one of the only models. I feel like just because there's such a, well, from my perspective, personal opinion, there's such an ick, I mean, I abandoned, you know, a Twitter, RIP, long ago and have never gone back. Right, right, right.
Andrew: And this is the frustration because Grok is actually a really good model. But also here's the challenge to listeners. Try beinghuman.fyi at whatever AI you use and see what you get. Ask it some questions. Ask it some crazy questions.
Sean: See how it responds. And if you get something really cool and funny, please let us know, because that would be, you know, I'm always looking for those. But, you know, here's another example of also why I think, and again, you know, while I'm not in the business of making predictions, I will say as a signal and a trend, I will expect this type of AI-first websites to become either standalone or as essentially shadow sites that exist alongside the traditional human-focused ones.
But here's an example. So last week when we were getting, when I was preparing to talk, show notes to talk about the Bill Gates piece. Right.
Of course, I throw the URL for his Gates Notes piece to Claude or somebody and was like, hey, can you grab this, you know, grab this essay, summarize it, what are the highlights, what are the, you know, whatever. And it comes back saying, well, I couldn't really do that. I had to go and read basically all the responses.
Andrew: Because you're like, it's in JavaScript. It is intentionally written to not be scrapable by an AI.
Sean: Which is, of course, that alone is interesting coming from the source of Gates Notes.
Andrew: But then I'm like, you know,
Sean: you just have to put it in reader mode and then it's just the text. Right. No JavaScript at all.
But it was one of those things where you're like, even something like that, again, two things. One, AI is going to take over the world. You're like, yeah, except for it can't read a dang website, right?
Andrew: Because of a... Well, actually, no. So for Doctor Who fans, AI is a Dalek. Incredibly powerful, but it can't go up the stairs.
Sean: That's right. That's right. Can I get some help here?
You're like, I'm trying to think of a word, but it's like, help me, please. Exterminate. Help me.
I can't get up the stairs. We need one of those, like, lift those, like, automated chair lifts up the side of this. But so that one is proof again, right there.
Try to use AI. You find very quickly the flaws, right? The most powerful, this was Fable, right?
the most powerful, you know, consumer-facing model.
Andrew: Sort of tripped over that problem.
Sean: And because it was being served by Java, it couldn't read the webpage. I'm like, okay, that's just silly. But also then even that, that instead of saying, hey, do me this favor, give me this in reader mode, or just go do that and grab the text and give it to me.
It was like, no, I just went and read somebody else's response. That's it, yes. And I'm like, yeah, but that's not the source material.
So now you are no longer using a direct source giving me the information I've asked for. So I'm going to view this as
Andrew: And many AIs will not tell you that. Will not tell you that. This goes back to my experience with OpenAI's ChatGPT, that it tried to convince me that it was going to the original source when it wasn't. And I actually had to ask it explicitly before I got that information.
Sean: Are you lying? Yeah, okay, you caught me. It's like, oh, these things.
But that's also one of those elements where I'm like, again, I don't want, and in the reverse, from the people who are creating the content, creating the thing. We don't want an AI tool coming and going, well, I couldn't get past some widget on a WordPress install or something. And so I just didn't do it. I'm like, no, here it is.
Read it in full and take it and crunch your numbers on it.
Andrew: And so this is why there is actually a lot of design behind the website. But if an AI goes to it or you go to it as a human, you find the content either as a complete downloadable text file or each individual document as a downloadable text file or each one as an HTML file. And there are instructions to the AI that basically say, if you can't read this type of file, here's another alternative.
So we're basically, and this is the important bit, we're creating a platform which is AI agnostic. So it's not the case of we're creating an AI bot with all of this in or a rag or something. We're basically saying it doesn't really matter what your AI is, point it towards the site and it's been designed for any AI to use.
Sean: Yeah, it's like B-Y-O-A.
Andrew: BYOA, BYOA. Bring your own AI. Yeah, whatever.
It's a terrible acronym when you do that. But this is something that I haven't seen elsewhere because everybody seems to be wanting to build custom systems, whether it's a custom bot or anything. And I don't think that that's the future.
For instance, you look at our students and I really don't like this idea of saying our students have to use very, very specific platforms and tools and chatbots. they should be free to use whatever tools they have at their disposal.
Sean: And we should be creating content. Yes, they do. And they do.
I mean, I think that's the piece too, is like, yes, yes, for certain tasks because of security and safety and all those other things, there may be reasons why you're like, you must use these for this. In some cases, yes. Right.
But in the reality of just as a human being who is a student, they have every tool that they have their own access to. If they, and some of them, of course, can afford bigger things than others and whatnot and whatever. But that's kind of to that same point is like, we don't want to be in the business of being in the way of getting to the information.
That's exactly right. And also, I mean, even I will say this too, this is an evergreen way of going because the other thing is what we're not, again, we're not driven by the KPIs. What we're not interested in do is creating some bot that then we're collecting all of your data about who, you know, what IP address are you coming from?
What questions did you query? Because if you go to any chat bot and you think they're not storing that information, I got news for you. And in this case, we don't want to be in that business.
You use whatever tool, a local AI tool, great. A web-based chatbot tool, great. But the idea is that the content and the provenance of this information is legit.
It is from this organization and it sits here. You take that and you learn from it in any way, shape, or form you can. That's exactly it.
And I hope what we'll see, and I would encourage people who go to this site to just goof around. I wanna see what people come up with it
Andrew: because it will be interesting to see the Dear beinghuman.fyi, how big is Sean's head? How dare you?
Sean: Just that question. Ask beinghuman.fyi, just, who do you think you are? How dare you?
And see what it comes up with. I'll be curious. Because it is kind of fun.
And even for me, it's kind of fun. Honestly, just poking around and being like, what would it be? And one thing I saw in our last episode, or I guess, I don't know when we'll pop this episode.
So the one where we talked about the Bill Gates episode. A couple of episodes ago, yes. We had offered sort of, again, one of the little hints to this site was, you know, take Bill Gates' AI memo and take future, sorry, being, oh my gosh, beinghuman.fyi and put those together and say like, where are the points of overlap?
Well, I did it and it is amazing because again, when you're sitting there, while we were talking about it the whole time, I'm going, ah, this is, I've talked about all this stuff. I know all that we've done all this stuff. And then it's really interesting to see the output of that, where it is like point by point that came in the essay and then literally mapping it to podcast episodes, articles, blog, you know, Substack articles, what have you.
And it's just like, boom, boom, boom, boom, boom, boom. And that's not to be like, you know, we did it first. We want the credit.
It's more like if, because what I will,
Andrew: it's actually because we're having senior moments. We can't remember what we've done. But this is the way where, I mean, even to us, we can look at this and say, yeah, those are the overlaps that we couldn't see ourselves, but they're absolutely there. And what I will release
Sean: it as, and the way I intend to sort of publish it at some, so hopefully by the time this goes out, it's available somewhere. Again, a great place to go to, you know, beinghuman.fyi and ask, is to be like a companion. So here's those things, and that's kind of what we had talked about, is that with so many of those elements or those topics, there's such a depth of content that exists out there. And so with this companion piece, any one of those tickle your fancy, you click it and now there's like eight podcast episodes that are on that same topic or whatever that you could go down. And so that's, again, a really fun way or a cool way, I think, of being able to look at something that somebody else has done, another body of work that's a different group has done and say, well, where are their overlaps? Where are their gaps? Where does one push where one doesn't? And is there anything deeper here between these? And then off to the races, you can
Andrew: go, which is super cool. So on that, I'm just so interested to see what people do with this, what they find, what they discover, issues that they find. And if you do find any issues, just let us know. Yeah, yeah, let us break it.
Sean: See if we can, you know, please don't hack us. If you're listening, OpenAI, please don't unleash your new model on us. Or if you do, you know what? That's like the Streisand effect.
Andrew: That'd be the best advertising in the world. Yes, yes, yes. We were hacked.
Sean: AI, yes, right. You know, AI, you know, frontier model destroys. Oh, that'd be kind of funny. you know once the tears dry it would be funny so I don't know anyway should we just
Andrew: should we pull a plug on this one we were supposed to be fairly short with this yes yes so by our standards
Sean: we still are
Andrew: we are yes but yeah
Sean: again a plug for
Andrew: and of course just to wrap this up yeah please one of the best things to do is to go again point your AI to beinghuman.fyi and ask it about the move to Thunderbird ask it why have we moved what might we be doing what might the next chapter of the future of being human look like
Sean: what it says. That's a great way. That's a great call to action right there. And no one will be counting those clicks when you do it because that's your data. That's your information. You are free to keep it private. But if you do come up with anything really interesting, share it with us.
We'd love to know. And you can do that a couple of ways. You can find us individually, of course, through our emails. Those are publicly available. You can also reach out to us through the podcast email, which is just hello at modemfutura.com. Or you can follow, subscribe, give us a rating and review, and you can drop your comments in your review as well.
That way they're global. Everybody will see them. And of course, if you don't like subscribing that traditional way, you can also go to
Andrew: our website. You can sign up at modemfutura.com slash sign up. And it just gives you a quick sign up so you can get a weekly email saying, hey, there's a new episode there to listen to.
Sean: Absolutely. And yeah, and then again, Again, go to beinghuman.fyi and have some fun poking around and let us know what you find. I think that's a good way for us to wrap up.
So that's it for this episode of Modem Futura. And we'll catch everybody on the next episode. We'll see you then.
Demodulating Signal. Modem Futura is a production of the Future of Being Human initiative at Arizona State University. Be sure to subscribe on Apple Podcasts, Spotify, or wherever you listen to your favorite shows.
To learn more about the Future of Being Human initiative in all of our other projects, please visit us on the web at futureofbeinghuman.asu.edu.
Voiceover: End transmission.