Marcio's AI keynote - The UX of AI, ChatGPT had unbundled, like the good old days of Google

Summary

Today I celebrate a unique moment in history - yet we are right there, it front of that simple and magical user interface, ChatGPT. And that reminds of of the good old days when Google changed everything, and made search to be simple.

The situation: the UX of ChatGPT, quite simple, no ads, straight to the point, we see it now. But we have seen that before, to a certain extent, like in the good of days of Google.

The complication? We add things to our great interfaces, incrementally. Everyday. But worse, we get used to it, our additions. We have seen with the search, with the browser, with mobile interfaces, with everything.

Let us reflect about "unbundling and bundling", about the marvelous moment when we unbundle things as we know it, when magical things simplify everything. And then too about that other moment, when we become part of that incremental movement.

The Google old days

You may or may not remember the good old days of Google, when it was simple and powerful? We might as well look at other good old days of search engines: let us go back in time to the Altavista days. Simple and powerful, it started like that:

But then, search portals changed, they got bloated:

And Google came, simple, just with one field there. And plain good results. Of course, it was not all due to that UX component. It wasn’t about making it a simple search field. No sir. For that simple search query to sustain itself it had to do a magical thing in the first place. It had to be Google, the way it was. Yet, that single and simple new UX, the search field, was back there:

Not so different, to some extent, from the ChatGPT UX we have now?

The situation, let us build

Now, we know of a situation, don't we? We know that all things which are good get to be, let us say "enhanced.” And there are good things from there, from enhancing things, we know. That enhancement is a natural part of evolution. And to be fair, product managers, they aren't to be sentenced. No, because they listen to us, us users.

So we can't go on without recognizing User Innovation. And I mean, MIT professor Eric von Hippel's User Innovation, and Eric is right: It's all around you, that is user innovation. And indeed, the user does, together, with producers. Product managers, they are followers. Thus from there, from the sum of all needs, we get to add things. Add and add, and add. And we know too what happens next. Or worse, we don't know because incremental adding can't be seen.

Let us talk about that: incremental adding, which can't be seen. Let us talk about the phenomenon of unbundling and bundling in connection with creative destruction. And let us reflect about when that portal opens up, like now. So let us celebrate the "UX of AI" using the case "The UX of ChatGPT", celebrate that known unique moment in time. It’s time to celebrate.

Incrementally adding

Here is a situation, back from the days when I was a student of Computer Science. Myself and my friend, Chris, we used to visit a lab of the geek professors, the folks using the Sun workstations. Those Sun workstations were special. And only top professors had access to them. But we could walk around the lab, and we had to do something. So we would go, and visit the lab, everyday. And we had the bright idea of rotating their display, 1 degree per day. The next day? Nothing different. And we kept going. And that was a big lesson we took, right there, that nerds, they or we nerds, they can't see. They can’t see subtitle movements.

This is what we get with these user interfaces from everything we do. And this may be so, of course, because Peter Drucker was right - in his little great book about self management. He was right in knowing that we want to be somebody. We are builders. So here I celebrate that drive, that need to build, that great thing that takes us there, in places indeed, but too to incrementally add, day after day. And sometimes too, up to a point when someone chimes in noticing something, and before we know it we say "a genius!" Well, we have to be somebody, right? So we scream out loud "you are a genius!" to celebrate the light because we just can’t see. It would be special, if we could, leave earth and come back, wouldn't it? But we are here, day after day, playing the fair game, living being the sum of all things, incrementally speaking.

Cases - Netscape to Firefox

Take the Netscape browser case, from the history of the web. Let us go back in time. It was just that web page navigator, written by Marc Andreessen and his colleagues. And it became, as history had properly archived, Netscape Communicator. And from that, a lot of things came, we know it.

But then, we noticed something too, we were adding things to it, on top of its value. The train is good, the train can take, let us put things there.

If we follow the history of Netscape, if we ride along through the AOL days, we can feel it: so many product managers and builders, doing it. Not alone, of course. I was there, to some extent, as an evangelist. And what is the job of the evangelist if not to hear the outside. So we heard, we heard a lot. And to be fair, it was a good job. But that was it, nevertheless, helping to put things on the wagon, helping build a product, bloated, full of nice things, on top, nice things, day by day, day after day.

And then one day, Mozilla was free from AOL. Or better, AOL let Mozilla go. And the mothers of Mozilla had assembled a new train, of course a lighter train. It had to be. And the young Blake, so of Mozilla, got caught on his mission too. And now he had permission. And he did it, as now we didn’t have to listen to so many voices, he could cut ties and cut through, access values. So Blake had indeed silenced everything, all the voices. He was onto something, to hear what would then appear to be a single voice. He put it that way: I will remove every single thing that my mother/grandmother does not understand. And he started to trim, or cut, or to trash. He started from the user interface, fine. So that was it, Blake reminding us that when we do that, that we can do that, and that we can do that from any angle you want to do it. Reminded us that it yields good things. Firefox was born, Mozilla reborn. Then the rest in history, and Mozilla by the Firefox days had too, as a by-product of that movement, awaken the other beasts. This is a case, an example.

Recap - bundle and unbundle

Let us talk about bundle and unbundle. A good early reflection about this idea came from Marc Andreesseen and Jim Barksdale. They nailed it, reflected on it, reflected about a phenomenon - the phenomenon that ties to creative destruction to some extent.

They came to that sentence at the end of an interview, when they were asked about the competition, with “How do you know that Microsoft isn’t just going to bundle a browser into their product?” And Jim told them, quite straight, that there are only two ways to make money: unbundling and bundling. Indeed, Microsoft did it, they did bundle "Internet Explorer", the browser, into their Windows operating system. And from there, after that, Netscape had no way, no way to escape. Netscape had to counter attack, let us say, to unbundle things, as Code Rush reminded us.

The user experience of building, or trying to learn, online

Now let's focus on building, now looking from the developer's view. Developers, web developers, producers, they are builders too. I want to celebrate the web developer's role. Of course, we know that they are builders and that they build together with others, say with product managers too. At the end of the day, as we know, they want to be somebody. And we know too that they all respond to that so-called user, to user needs.

But we are users too. I am a developer, a web developer myself. So I am just a user when it is time to learn. And we do learn every day. And how do we learn? We learn with examples, we learn with conversations, we learn by doing it. And we learn by using search engines, and forums, and new things every day, and tools. With all that, incrementally, we learn with whatever comes our way. This is so because, developers know, because the web always served, first, web people.

Web people, the first to feel the pain

But from that, we know too, they are generally the first to experience, or to reflect about web trouble. Let us consider web trouble. One case? Ads. We noticed first, too many ads, didn’t we? And we did something, didn't we? We killed ads, with ad blocking! I won't forget the days when my good friend Fabricio Zuardi made BannerBlind, which was ranked 18th on MozDev.org. Note: MozDev.org was by the 2000s a prelude to "apps marketplaces".

They first to suffer, incrementally speaking

But with that, as we spoke, them builders or we builders, they are the first to struggle too. But it's not easy to leave that bubble of struggle, because it’s not easy to see. One thing, we considered, that we are trapped listening or pressured by listening to so many voices, ties.

So in terms of the learning experience of developers, that became bundled too, with too many layers. Layers and layers. Too many ads, too many comments, wrong. Too many arbitrary thoughts, too many blinking things on the display. They got bundled too.

That is what we got, what they got, the experience of a young learner, learning in new or fragmented ways. So here it is a good time for us to celebrate ideas from Stolen Focus:

“Take care of what technologies you use, because your consciousness will, over time, come to be shaped like those technologies. “ Hari, Johann.

But here is a complication: developers, or builders, think that they are immune to that. Oh, I am a good reader, I am not part of that. I can separate ads from content, I can be good at following signals from noise. But how can they know? If the web increments and increments in front of them, everyday. When the whole web changes, its angle, their displays. If it is the web that gives them how to see, therefore what to see. What you see is what we get, isn’t it? And plus, we got paid to build that way, to see that way.

Unable to see, something bloated when learning

So too many developers started to learn by checking questions and answers, on questions and answers sites. By trying to pick answers by using the magical tools, shortcuts, in front of them. Ads, skips. Reading, diagonal. From a tools/infrastructure angle, taking the blocks others were giving.

So the learning cycle changed its angle. Now, by the lean terms. And with that, with an attitude which their bosses clap, giving them badges, stars:

You are lean, and mean, you are a fighter, you are a hero. Lean mean means a developer machine. You slide on rails, you do outstanding, standing work. You break first things, you are not afraid. This is collaboration, this is our culture.

To do something, feels wrong

But when if these user interfaces, when it seems that the UI of all these things fail, when that thought crosses their minds, when it happens, when it almost happens, they can't express it:

They can't say "hey, this workflow is broken." Oh no, that would be about expressing either a sign of weakness or that they want to stop the train, the lean train, that lean mean and happy train. Hey, are you trying to stop the train? You are not being a colleague, my friend.

Many sentences came to calm down those weird thoughts, it’s been a while. Every single country, or team, has their culture to silence the one feeling weird about keep going, about keep cranking. And hey, come back to reality my friend, because it’s working, isn't it?

And on top of that, we have toppings too, like foosball tables. Perks if you will. So we think that foosball tables and cokes aren’t connected with culture. Well, they aren’t, we did our homework. They are perks, indeed. In theory. But it may they are as well toppings for the day, for when that icecream is not sweet enough.

So take it, intake, and from that we keep on playing the game, using the same channels we know, working to fix our sites, our products, our learning experience. We then dive into debugging, again navigating the same sources, by checking examples from sites, documentation channels, forums, and more.

And the worse of all, unable to judge

So when the developer almost sees something new, like that there is something wrong with the display, he or she comes to another conclusion too:

That "I am the stupid one" or "I am wrong" or "this thing is right and I have to try again, to learn, to focus.” Here I would like to point out an amazing story from Executive Communications documented by Silicon Valley investor Michael Dearing. Michael brought the story of Barbara Minto that for me revolutionized a way to look at communications. According to Michael, Dearing, who worked at Walt Disney, Michael Eisner the then-CEO had a deep "religious conviction that Barbara Minto had cracked the code on executive communication." But Michael told us too something else from Barbara, before she had cracked the code of communications. The thing was that - while in Harvard - she looked at that tough calculus book and had an insight. She thought that the book was badly written. That was something no one at the time could see, so called smart folks from Harvard.

So here I remind us of the days when we almost smell trouble but we get back to trying hard to be somebody, judging us as the ones not doing enough in the first place; setting us back to that position, the developer or doer or builder, the one who plays the game, before foosball and cokes. Is our way to learn broken? No way. Too many ads? Well, you should focus. Too many sentences in front of you? Deal with it, you are a hero. Or try to be. Come on, it’s an ad. That is it.

The UX of chatgpt

Then we get something simpler, again. A conversational user experience, no ads so far, and quite to the point. Something new that for many sets them to calmer. Or learning.

References

• Mangelsdorf, M. E. (2011, September 21). The user innovation revolution. MIT Sloan Management Review. https://sloanreview.mit.edu/article/the-user-innovation-revolution/​

• Wikipedia contributors. (n.d.). Blake Ross. Wikipedia, The Free Encyclopedia. Retrieved April 8, 2025, from https://en.wikipedia.org/wiki/Blake_Ross​

• Fox, J. (2014, June 24). How to succeed in business by bundling – and unbundling. Harvard Business Review. https://hbr.org/2014/06/how-to-succeed-in-business-by-bundling-and-unbundling​

• Winton, D. (Director). (2000). Code Rush [Film]. PBS Home Video; Winton/duPont Films.​

• Hari, J. (2022). Stolen focus: Why you can't pay attention—and how to think deeply again. Crown.​

• Dearing, M. (n.d.). Executive communication with Harrison Metal [Video]. Heavybit. https://www.heavybit.com/library/video/executive-communication​

• Minto, B. (n.d.). MECE: I invented it, so I get to say how to pronounce it. McKinsey & Company. https://www.mckinsey.com/alumni/news-and-events/global-news/alumni-news/barbara-minto-mece-i-invented-it-so-i-get-to-say-how-to-pronounce-it