# Mozilla CTO on Open AI Strategy

**Podcast:** alphalist.CTO Podcast - For CTOs and Technical Leaders
**Published:** 2026-10-15

## Transcript

Hello friends, this is the Alphalist podcast.
I am your host, Tobi.
The goal of the Alphalist podcast is to empower CTOs with the info and insight they need to make the best decisions for their company.
We do this by hosting top thought leaders and picking their brains for insights into technical leadership and tech trends.
If you believe in the power of accumulated knowledge to accelerate growth, make sure to subscribe to this podcast.
Plus, if you're an experienced CTO, you will love the discussion happening in our Slack space where over 600 CTOs are sharing insights or visit one of our events.
Just go to alphalist.com to apply.
Welcome to the Alphalist podcast.
I'm your host, Tobi.
And my guest today wrote a sentence I cannot get out of my head.
He said he doesn't want seven AJIs, one for each of the big techs.
He wants seven billion.
He scaled Twitter.
He ran Uber Self-Driving Center.
He was the first CTO that the Democratic National Committee ever had.
And he also did a tech podcast that charted way higher than mine ever did.
So I'm a bit nervous.
Welcome, Rafi.
Rafi, welcome.
You're the CTO of Mozilla right now, which is great.
And really, I think we can learn a lot today.
Well, Toby, thank you so much for having me.
My first...
Like, let's say, famous question.
I ask every guest, what is your personal nerd path?
So when did computers actually enter your life and why did they stay?
Yeah, no, it's a great question.
So my grandfather, so, you know, I'm half Armenian, half Filipino.
And my mother's father actually started one of the first radio stations in the Philippines, like back in the 1950s.
And what he did—so he was the tinkerer of the family.
What he did is he set up a radio station in a small village called Kagehandioro.
It was right on the water next to the mountains in the north.
What he did is he allowed all the farmers who would come down to that radio station, to the farmer's market every day, to get on the radio to broadcast that I'm coming home so that they could send a message and their families could hear them back up in the hills before they started the multi-hour journey home.
So I think about that a lot because I think in a lot of ways my grandfather is probably what started...
my path on tinkering and computers and stuff like that.
He bought me my first computer.
You know, I remember when he bought an Apple clone in Hong Kong when he went to visit there from Manila.
It was called the Pineapple.
He bought me my first x86 computer.
He bought me my first CD-ROM drive.
It had an encyclopedia on it.
So in a lot of ways, it was that.
And then I would, you know, I would go to the library.
And, you know, when I was a young child, I would borrow books of like computer basic programming and my typing skills weren't so great.
So I'd recruit my mom to actually like type in all the programs so I can try to run them and try to modify them and stuff like that.
So, yeah, I think it's from all that.
And I guess there's some gaming involved also in a certain stage, or is that...
Actually, I never truly...
I mean, I had every single Nintendo system because I always wanted to get into gaming.
My grandmother, my grandfather's wife, bought me my first video game system.
She actually bought me a Super Famicom back in the day from Hong Kong.
So, like, you know, all my first video games were actually, like, you know, Donkey Kong in Japanese or things like that.
But...
You know, I wanted to play video games.
Like, my mother bought me a Nintendo.
I bought myself a Super Nintendo.
My kids now have Nintendo Switches.
But video games never really got to me.
The video game I do play constantly is Kerbal Space Program, because then I can, like, launch small rocket ships into space.
But, like, besides from that, like, I never really truly got into gaming.
Okay, okay.
Interesting, because most people kind of had that as kind of a way of barrier, like an entrance barrier.
No, I mean, because I went to MIT as an undergraduate, you know, with the full intent of becoming an architect.
And so I took a whole bunch of architecture classes.
But I also, by luck, happened to register for my freshman seminar at the MIT Media Lab.
And the MIT Media Lab is sort of like what really taught me like the power of computers because the amazing thing about the Media Lab is that they just live 10 years in the future.
So you walk in the door and you have all these technologies that...
regular people don't have, but they probably will in 10 years.
So like when I first got there as an undergraduate, I was assigned this SGI workstation and like no one had 3D on their desktop.
So like I had 3D on my desktop and we were like playing with what we could do in 3D.
And, you know, a couple of years later, we all had 3D printers at the Media Lab.
That was before anyone even heard about 3D printers.
And now, you know, there's a 3D printer sitting next to me here in my flat.
So I think it was like, it's those type of formative years.
So it wasn't really gaming.
It was like, Just this idea of just like being thrust into the future.
And then what can you do?
What can you backport to the present?
Now, I'm curious what happens in the Media Lab today.
Like, what is there what I don't have?
No, they're doing a lot.
So I'm actually on the board of the Media Lab.
I'm very lucky.
in that way, they've been doing a lot of biological computation.
So one of these amazing professors is working on small little circuitry that can cross the blood-brain barrier.
And then you can actually do brain...
not augmentation, but repair brain damage by these small little circuits that are crossing a blood-bain barrier and you activate them using magnetic fields on the brain.
There's another person, Hugh Hare, who builds these incredible prosthetic limbs that like you can, like, you know, most people train a prosthetic limb by using AI models to figure out which way the ankle should bend and stuff like that.
His answer was like, what if we just jack it into the nervous system so your brain can just control it kind of thing.
So like, that's the kind of world that the Media Lab is living in these days.
You should still go visit.
That's great.
Okay, cool.
Yeah, then let's hope they let me in.
And how did you then, like, I mean, you have a very like exciting history of companies you've been CTO at or participating in leadership, et cetera, like Twitter, Uber, DNC, Emerson.
Like, can you tell me which connective tissue led to Mozilla now?
Like, why are you with Mozilla?
I mean, like, in a lot of ways, you know, I can tell you both the short and the long arc.
The short arc is, you know, I was at Emerson Collective, and I was thinking a lot about how technology impacts society, how technology impacts affected populations.
And then I was thinking a lot about...
what surveillance technology could do to people, how data extraction was actually happening, how all these pieces of technology were actually causing potentially philanthropic problems, like things that philanthropy might have to go and try to fix in a community years later.
We saw this in mobility.
Again, we saw this in surveillance.
We saw this in data harvesting.
And we see this just in education in general.
Most of these communities don't get access to technology first, so they're always perpetually behind.
But then I started watching what AI was doing across all these different communities as well.
What disinformation was doing, what disproportionate access was doing, and things like that.
And I was starting to become of the opinion that AI was going to be, you know, this was years ago, AI was going to be this transformative force that if not distributed equally, or if people didn't have equal access to that, then that could become one of the big problems of our time.
And so I was coincidentally already on the board of Mozilla, but I felt that my life's work should become actually doing this work, not just governing the work.
And so like a lot of my work at Mozilla these days is thinking about...
How do we have access to all this new intelligence systems, not just have it solely concentrated in the places of a few big companies, but how can we truly democratize?
How can we truly diffuse all that technology in the way that the web has done?
But can we do that again in the AI world?
So maybe that's like the short way to look at it.
But the long way to look at it is I've always been interested in this intersection of people and technology.
Like my work at Twitter.
You know, I joined Twitter because we naively thought that things like Twitter could flatten the world.
Like anyone can talk to anyone, which at the time we thought was a good idea, might have been a bad idea in the grand scheme of things.
The reason I joined Uber self-driving was, or led it for a while, was because if you have self-driving cars, you could fundamentally change the architecture of cities.
Like right now, 20% of major cities is parking.
Like that's how the landscape is used.
What if that could change all of a sudden?
The DNC was about...
people participating in democracy.
So Mozilla, I think a lot about how do people participate in this new revolution of technology coming down the pike?
And not just this one.
Whatever is the next one, whether it be quantum computers or something, how can people participate in it?
Okay, then we talk about your view on AJI later.
Let's start maybe, or let's go a bit deeper on your personal AI journey.
A quick, quick check that I want to ask every guest from now on, because it's quite interesting, and everyone's experimenting like hell right now.
How does your personal harness look like?
What do you do personally each day?
And yeah, let's start with that.
Oh, I mean, my life is in a constant flux right now.
I'm trying new things all the time.
My primary harness that I'm using is an Hermes agent that I named Zora that has access to all my things.
And I like that better than using, say, like...
dots or one or grok bot or whatever, because it means that I can run it all effectively locally.
So I'm running that I have a DGX spark in one of my closets that's running the model that I'm running for it.
So I'm trying to do everything as locally as I possibly can, because like, that's the experiment that I want to try to push.
And every once in a while, you know, I let it have access to something like a fable or an opus when, you know, when a GLM just can't pull it off.
But most a day, maybe 80%, 85% of it, I'd like maintain low.
locally on that harness that's running in my closet.
But for my development environment, I have like, you know, you can't see my desktop.
I have like five different terminals open right now.
And one of the biggest experiments I've been running, so I'm a big, both Claude Code, but also OpenCode user.
And one of the experiments I've been doing with Claude Code, although the Anthropic people are probably not going to like this, is like, I've been writing code to try to manage multiple subscription accounts because I'm actually, I'm really concerned that we're going to get into a world of like, subscription pricing is going to die in a post-API world.
We're all going to move to token pricing.
So I'm trying to maximize the amount of use I can get out of a subscription.
So I have like three different subscriptions that are automatically rotating behind the scenes.
So like none of my agents actually know it.
And so like I've been doing that for a lot of my work.
So you don't use the original Claude harness, but you use something like OpenCode or so and then switch the models in the background?
So I use OpenCode as an experiment for like how I can be using both Kimi and GLM from just to my end, like, you know, Jed.
stuff like that for most of my day-to-day.
But for like when I actually need to really, truly write code that I want to know that's good and it's going to work for a while, then I'm using Claude with Opus.
But again, I have like 12 different projects going at any given time.
So even the like 20x max subscription is not enough.
So I just have three 20x max subscriptions and they're all just sort of like rotating in the background behind it.
So I have like a meta harness that I'm running Claude inside of to actually manage that.
And your meta harness basically runs multiple tabs then?
Yeah, so it's actually running multiple subscription accounts and it's trying to keep track of the rate limits at any given time so it can switch you between whichever one has the most rate limits open.
And then as part of this, and I should open source most of this, and I will.
Like that one I call Rota, which can rotate all the subscriptions.
But then I have another piece of code that I call the Corral, so I can actually put a true sound.
sandbox around Claude code.
So there, I actually control network access, file system access.
We actually cheroot it into a jail.
So then that way I can be explicitly sure that nothing is jumping outside the boundary without my permission kind of thing.
So I'm really curious about all this.
And then you use YOLO mode inside or what?
Yeah, exactly.
But then I can run auto mode because I don't care because it's going to be stuck inside the jail and I'm not worried about it.
Yeah, yeah, yeah.
Okay.
So honest feedback on Hermes, I try to do it as well for a while.
I have a 128 gig unified RAM computer at home and I think I run GLM right now.
And I don't know, for me, the quality of the Frontier models is still so much better that I think it's, I really want this to become not only a trend, but the future.
But I think it's like, for me, it still takes a while.
And then with Hermes, I don't know, I always fall back to Cloud Code because I realize, okay, you don't really need it.
Like you have, I mean, I also have this, like I have a tool called Agent Deck, a very simple open source tool, which basically orchestrates all my sessions and it has a pattern, a super tool they call Conductor, which can watch multiple sessions, et cetera.
It's kind of helpful.
And I realized I don't need more.
I don't even need that, probably.
So I have like 10 tabs open.
And then what do you productively do with Hermes then?
I mean, Hermes is basically running most of my calendaring, meeting briefings, all my scheduled jobs that need to sort of like run in the background so I can wake up and just get going with my day.
I take your point that for interactive sessions, I find it a little slow, and especially because I'm trying not to run it on the Frontier models.
It just...
it takes more turns in order for it to like finish its job.
So that's why in a lot of ways, like all my briefings, like if you go into my calendar, if I could show, I'll show you a screenshot one of these days, but all my calendar events has like a corresponding calendar event next week, which has like, here's everything you know about this person.
Here's every, all the summaries about the email.
So all that runs while I'm sleeping, which is fine because like that, I don't need to be in an interactive session.
And so it's that kind of stuff that Hermes is just like running for me all the time.
Okay, okay, understood.
Yeah, I mean, for me, it's kind of the same.
I also have one running and it kind of has access to my WhatsApp, etc.
And it summarizes stuff that happens also in my family life, like, hey, you need to arrange this for your kids or so.
It's kind of helpful, but I also extracted this into kind of a family asana, let's say, that also has an inbox where all the stuff is arranged, etc., which is explicitly written.
And maybe I'm a bit of a control freak still, so I want to...
kind of have control about this and understand basically how is it built or not how is it built but like I want it to be repeatable.
Yeah, well, I view this as sort of like the same way that I talk to companies who use AI of just like there's an experimentation phase, then there's the deployment phase.
So I spend a lot of time in this experimentation phase of just like, I have this idea.
Let's just tell it to Hermes and let's see what happens.
We'll tweak it.
We'll do all the things.
And then at some point of just like something might graduate, I'm just like, okay.
Now I know enough.
Like now we can just build it for real and just let it run in the background.
So I've done that for a few different tasks of like something that used to be in our maze.
I'm just like, no, we can now make a real tool.
Like I now understand what this needs to be.
It's like, you know, in some ways, like the product discovery life cycle has changed instead of like me sitting down and like really thinking about it.
I'm just like, we'll just work rapidly, prototype and live with it for a few days.
And then that'll get me a better understanding of the thing I need to extract.
Yeah, that makes sense.
That makes sense.
And what is your, like, I have friends in my office, like every day I come in, I shout loud, like, hey, can you believe this?
This is what just happened, right?
Do you have like any aha moment for you that recently happened where you said like, hey, this can't be true.
Like this is magic.
So let's go back to when I was talking about 3D printers.
So my wife is a big hiker.
She takes me and my family hiking all the time.
You know, we recently scaled the largest, I think it's technically a mountain in Wales.
I mean, it's small compared to other countries.
And so we had the GPS track of it on my Apple Watch.
So I gave Claude Code access to the Apple Watch, gave it the GPS track and told it I want like a model.
that has like the terrain and the map of like where we hiked.
And then a few minutes later, it spit out a 3DMM model.
I upload it to my bamboo printer.
A day later, I have a physical thing sitting here, which is sort of like the app.
I'm just like, now we've jumped from just writing code to like real things are showing up into the world.
So for me, for me, that's it.
Okay, that really is magic, like this intersection between physical world and a prompt, right?
I used to build self-driving cars, but now because I just did that, I'm getting very excited to be like, my kids should start doing 3D printed robotics because we can now just from a prompt create a motion device that we can then program and just do it.
Yeah, this is magic.
I also had a funny moment yesterday, I want to tell you.
I have a coffee grinder, an Italian coffee grinder.
And I don't think that Italians or Italian coffee companies should do software.
But it has a display.
For whatever reason, it has a display to tap between like one and two shots, right?
Like espresso.
And I discovered it has a USB port, so you can firmware update it.
what's wrong with it?
The clock is broken.
Whenever you turn it off and on, the clock is like, I don't know, two hours late or so.
And I don't know, I thought like they must have fixed this through some firmware update.
I can just flash via USB.
There are firmware updates.
Nothing, right?
For years.
And yesterday, I just downloaded that firmware again and gave it to Claude.
Fix it.
Yeah, what's this?
What's this?
And it just, it figured out everything, which chip it is, how it works, etc.
And then I said, okay, now I don't feel YOLO enough to kind of flash it, right?
It fixed the clock problem, seemingly.
Can you build me an emulator?
And guess what happened?
It's amazing.
Yeah, yeah, this is really amazing.
Like you can, in your browser, it runs an emulator.
That's so cool.
Yeah, it says when it stops the grinder, when the grinder's on.
I mean, it's nothing special in a way, right?
No, my 10-year-old...
has been vibe coding video games.
So if you go to like cute and stinky dot productions, there's all these video games he's just been vibe coding.
But one of them, in a similar way, he wanted to do, he wanted to vibe code a Tamagotchi.
Like you remember back in the day when you have those virtual life pets.
So he too, like he has a Tamagotchi simulator that he uploads firmware to and then it sort of like changes all on the web.
It's so wild.
That's really wild.
Yeah.
Like that, that, um, I think this is like a window of opportunity now until all the vendors really lock down their systems, right?
Because, yeah, I mean, it won't stay secret, right?
Like you can hack everything, basically.
Let's see.
Let's see.
And what do you think is the effect on software from your perspective?
I mean, you also produce software with Mozilla.
I mean, you have like...
kind of, I would say, also very special distribution mode, right?
But what's your view on the software development lifecycle at first and on software distribution?
No, I've been actually worried a lot about the software development lifecycle because I think that like if you just think about the social aspect of writing code, right?
I used to run large engineering teams like the Twitter engineering team was 600 people.
The Uber engineering team was a thousand people working on the car.
And so like a lot of it are social contracts.
Like writing code is not just a matter of hitting things on a keyboard.
It's like coordinating with another team or figuring out a requirement or stuff like that.
All these tools have now increased the power of a single individual, and we haven't really figured out what that social contract can look like.
Which, you know, if you start looking at all these open source pieces of software that's been vibe-coded...
their product skills are crap.
It's because they've dumped the kitchen sink of features into them.
They're like, I have 12 different command line flags that you can now toggle on and off.
And so that becomes software engineering in this world of agent decoding has become like a very solo thing.
Like a single person now is like doing all this.
what in my mind is craziness.
And what we need to figure out is just how do teams work together again in this agentic model?
Like, what does an SDLC really look like when you have multiple people who are agentic coding?
And I think, honestly, the tools are not there.
Like, for example...
I don't think, you know, we all rely on Git as the mechanism if you're working on Teams.
I don't think Git's the right tool anymore in this agentic world because like Git is about committing like keystrokes or like committing like bytes, not the compiled code.
But these days, no one's typing those bytes anymore.
They're typing prompts.
So like, what's the tool that we need to share the prompts across all the teams?
So like, I'm actually really, I think we're going through this moment where like, we're just generating shit when it comes to software.
And we need to sort of like, reform our tooling, reform our processes, reform the social contract, so we can start writing code again properly.
We're seeing this in the open source ecosystem.
Like, you know, for whatever reason, open sourceness is still staying true, right?
Like, you don't like...
in the world that people don't need to collaborate anymore to write code, we're still sharing all this stuff.
So I still have a glimmer of hope that we can make it work, but we got to, we got, like, I think a lot of people claim that in this world, we're not going to need engineering managers anymore, not going to need product managers anymore.
I think the exact opposite is true.
We might actually need more engineering managers and more product managers until we settle down the social contract again.
I think there's an infinite appetite for software as well, right?
I mean, there are so many people that wanted to have software, that weren't able to have software, that wanted to build something, that weren't able to afford the developers, etc.
This is all changing now, so I hope it's also like the results.
But to that point… Like, you know, a dentist can now vibe code their own website.
But like that website is going to have like five different security vulnerabilities.
And like all my PII is going to get hacked out of it kind of thing.
So like, well, at least right now, right?
I mean, the knowledge gap, I think is closing, is shrinking, right?
Like if you look at Opus 5.5 now, it's really like very good quality.
So we need better tools.
Like the tools need to catch up into like regular people doing this.
The tools need to catch up so that team.
can work together on it.
Yeah, I think that's the state of where we are.
But coming back to then teams of software engineers, I agree that like there's no, it became in a way a single player environment again, right?
But I mean, for me, I in a way like it as well, to be honest.
Like when I, I think the best time in software engineering for me was when I was working with three great individuals that were just super productive, that didn't need any tickets, that didn't need any requirement engineering in a deeper way, but it was just about letting it flow and having fun.
And I think this just returned in a way.
I think the challenge is really to find the right three individuals to work with, right?
Well, maybe you and I should have five good together.
Probably.
And the question is, what do the others do now, right?
And yeah, I think that one big issue is that AI is in a way an accelerator of the good and bad, right?
So it also leads to dramatic...
No, I mean, experience and taste matters.
And I think that, like, experience and taste don't come from tools just rapidly accelerating.
They come from, like, the battle wounds that we all have of, like, I tried that before and it didn't work.
Or, like, that thing is really ugly.
I've seen 30 versions of that ugliness, but this is the one good one.
So experience and taste still matters.
And, like, we need to figure out how to, like, spread experience and taste.
Because now the tools are making us have this superpower kind of thing.
I think it's also the ability to change flight level, right?
And I think this is something to also teach and spread, right?
That you kind of, I don't know, whenever you decompile the coffee grinder firmware, you know that there's such a thing as emulators, right?
Like you want an emulator.
To actually ask for the right things and have the right wishes is super important, right?
From my perspective, I also teach how to think bigger in a way, which is an issue of software engineers that used to receive basically linear tickets for a while now.
Everyone has to become the one who writes the tickets, right?
And this is a fundamental issue.
And you need to know when's the right tool for the right thing.
Sometimes you need a screwdriver, sometimes you need a jackhammer, but they're not interchangeable.
So you just need to know.
Yeah, yeah.
But let's talk Mozilla a bit, right?
So what is the big plan to make it like billions of AJIs instead of five or seven or however big techs we're seeing in the near future?
Like what's the plan?
Yeah, I mean, so we started the year, or we started actually the middle of last year, trying really hard to boot up the open source AI ecosystem with the theory that when more developers are using open source AI, then it'll have a proliferation of both different models, but also different applications, etc.
Like, you know, if we look at this concentrated world we live in right now, there are two, maybe five big players when it comes to...
big AI and the best products that we've come out of, except maybe recently with news and stuff like that, were like, text boxes.
So like, so like, I want to live in a world where we have like a proliferation of like amazing applications, similarly to how the web worked.
Like the web, no one predicted that you could have a Twitter, no one predicted you could have a TikTok, but someone was allowed to just do it.
And then sort of then took over what that web looks like.
So I'm hoping the same thing could happen with AI if we can get more technology, more diverse technology into the hands of as many developers as possible.
possible.
So that's where we started last year.
And the world has caught up to us.
I'm not going to claim any credit for it, but the world has caught up to it.
Like, if you look at the conversations on X, if you look at the conversations on LinkedIn, if you look at conversations on Reddit.
Developers worldwide are talking about open models.
And in fact, like a lot of the data shows that developers are using open models right now.
So the question now becomes like, how do we continue that proliferation?
And I think the next step is to not just convince the developers, but you have to convince their bosses.
So like right now, like people like you and me, like we can make the honest trade that if I want to write some code, I could use cloud code or I could use open code.
But when we go to the business level as the next level, some of their bosses, some of their CEOs are still kind of skeptical.
Like, who am I supposed to sue?
Like, what is the true cost of doing this?
Because it's not just like a credit card and token cost.
Like, how's that all going to go work?
And so like a lot of what Mozilla's work right now is, is if you assume that the developers are starting to talk, how do we...
basically soften the ground so that their CIOs, CTOs, CEOs are going to be okay with it as well.
So we're spending a lot of time talking to the equivalent of the Fortune 500.
We're talking a lot to governments for investment purposes to try to explain to them and try to soften the ground to make it possible for open source, open weight models to be deployed and bring the technology around that to make it easier and more possible.
But isn't also bottom-up a good strategy here?
I mean, we were talking about open code.
Should you just acquire open code?
So the beauty of open source is I don't think I need to acquire them.
I think I need to build an environment where they can thrive in the grand scheme of things.
So I think a lot of what Mozilla did when we came to the web is we didn't necessarily need to take the entire web.
In fact, we didn't.
And the fact that our usage share has diminished, honestly, but we've still won and keep a web to be open.
Because the existence of a Firefox, because of the existence of a Mozilla, the web is run on open standards and no single company can control it.
Like when Google decides to make a proposal to change the way that something works, WebMCP, etc., Mozilla can stand up and be like, that is not...
That seems to be biasing the Google way that doesn't bias toward making the web open.
So we can force those conversations to happen.
And so I don't actually need all the power to be in a single place.
We just need to have lots of voices that when bandied together can push back.
and make sure that this technology is accessible and open to everyone.
So we could go acquire an open code, but I don't think I need to.
I just need to make sure there is an open code.
I need to make sure there is an open wait model.
I think to make sure there is an open harness that can be used.
And sure, we can have a rebel alliance where we all talk to each other, make sure we're coordinated, but we don't need to all be under the same company in order for that to be successful.
But a little...
provocation.
I mean, you're also backed by Google big time, right?
I mean, Google pays a lot.
Yeah, Google, the Google revenue stream does.
does pay for a lot of it right now.
But in a lot of ways, like that percentage is constantly going down.
Like the nice thing about Mozilla is that we are able to now experiment and try lots of different revenue sources.
So like, you know, the Thunderbird team has a different revenue source than the Firefox team has.
We started a Mozilla data collective company, which has a different revenue source.
So like we're diversifying ourselves as well.
I would also say it's also a big chance that there are no...
other big players in the market, right?
So as long as this user interface, the browser exists, that you can embed others as well, right?
Probably.
100%.
And I think there's also a future.
And I think there's also a future for what does a browser look like in the agentic world is actually a pretty important thing.
In fact, the vast majority of traffic we're starting to see is from agents.
And so using something like a Firefox from an agent actually opens up a lot of opportunities as well.
And how do you concretely then make sure as Mozilla that an open code exists, for example, to stay with that example?
Like what are you doing and what are you doing for the open-weight models?
Yeah, so we published something that we called the State of Open Source AI Report a few months ago.
It's a living document.
It's at stateofopensource.ai.
And every few months we update it to have the latest set of information of how the open-source AI ecosystem, open-weight ecosystem is going.
And so in a lot of those, we've mapped out what the space looks like.
We've mapped out what we think that open source AI stack looks like, mostly for the inference side, not the training side, but we'll get to training at some point.
And so since we have that now, we can do both targeted investments, but we can also direct others.
So the nice thing about being in Mozilla is that governments call us up.
Companies call us up all the time.
And then when they ask us what we should go do, we're able to then help them and direct them.
The government of Canada was asking how we can proliferate, how we can compete as a middle power country in this new AI world.
And so we pointed them toward an opportunity to go build an open stack in collaboration with Mila.
When another company contacts us, we can talk to them about, you should do an investment in the company XYZ because that will both help you and keep the ecosystem open.
So we can play almost like...
I wouldn't say traffic cop, but maybe a router when it comes to investment purposes of just like what does that map look like and make sure that money is sort of being distributed across that map.
It sounds a bit like a foundation or...
I mean, Mozilla at its top is a US nonprofit.
And so like we do have subsidiaries, which are for-profit companies, a corporation being one of them.
But, you know, we're driven by a mission to keep the web, the internet as open as we possibly can.
We have to be sustainable to do it, which is why we need a revenue stream and which why we're very grateful for like the Google revenue stream as well as the other ones.
But no, we are a nonprofit.
Like we need to be both...
mission and money-driven.
And are you then also involved with the core products that you have at hand?
Like, are you involved with Mozilla or with Firefox?
Yeah.
I mean, my job as the CTO is to think about the big overall technological arc that we need to operate in.
Each one of the companies has their own engineering team.
So, you know, Bobby is the CTO of the Firefox team, and he and I work closely together.
It's because of Bobby's team that the web is open.
And so like now we're booting up other teams to like work on making sure the data sets are open, to make sure developer tools are open, to make sure we're doing the right investments.
We have a whole Mozilla.vc team that just does investments into the open source business ecosystem.
So I think about across all the different levers, how do we pivot us in this moment toward AI?
But also how do we look around the corner of like what might be coming down the road next?
So for example, the foundation has been doing a lot of the work for me to think about what would open quantum computers look like.
And so we can start thinking about like those longer term arcs as well.
Oh, okay.
Okay.
Interesting.
And is there also a plan to also boot up new products in that AI era and to actually build on models, et cetera?
Well, we probably won't build a model, but we've invested into model building companies.
So like, you know, we were investors in Huggy Face.
There's a woman named Sarah Hooker in San Francisco who started a company called Adaption Labs.
And what they do is they build really easy recipes to rapidly create.
fine-tuned, custom, small, single-purpose models for your domain.
And then what they open source is the recipes in order to make those models.
So we probably won't ourselves build our own frontier model or open model.
But we're in conversations with the people that do them.
So we talk with the Apertus team all the time here in Europe.
But I think about the entirety of the stack.
So if we think about it on the application layer, The team behind Thunderbird, our email client, has built something they call Thunderbolt, which is basically their...
their chat client, which has access to your context, has access to your calendar, has access to your history.
So then you can then chat with all their personal information in some way.
So we'll be releasing things like that.
Firefox has Smart Window that's been slowly rolling out to provide AI-like functionality.
At user's choice, they can select the model, they can not use it, they can use it, but so they can have AI functionality as well into Firefox.
So we are building more and more consumer products.
We're getting there.
We have a lot of work to do.
We're not there yet by any means.
What is your honest take on the future of the UI?
I mean, if I look into my cloud sessions or my chats, I sometimes think, should I deinstall my email client?
Honestly, how does that look like from your angle?
also like raise the necessity to also build an alternative to those UIs, right?
Or to those simple interfaces.
Is that, is there also a secret plan or not so secret plan to do that?
Well, I can say that from my own personal usage, like we talked about the Hermes agent, we talked about all the stuff that I do.
I can tell you that my personal browser usage is down.
Like I use a browser less every single day because every surface is now.
computationally smart in some way, so I don't need the browser to do it.
But the browser still maintains a core place in my everyday workflow, whether it be for infotainment purposes or watching movies or playing games or whatever it might be.
And the browser has its place when it comes to agents still need a way to interact with the web.
And if most of the web is still built around these web-like technologies, the browser is one of the best interface conversion layers that they can go do.
But I hear you that like, well, we need to be thinking about this things like what I mentioned, Thunderbolt.
I was just like, what would new access to that kind of information look like?
How do we make all surfaces smart?
in some way, shape, or form again.
So we're using our venture team to do a lot of that exploration for us.
So we've invested in companies that are doing generative UI interfaces, like imagining a world where HTML is not sent to you by a server anymore, but instead just information is coming and my client needs to assemble it and render it for me somehow.
So the venture team is doing a whole bunch of this exploration of what would a personal GPU look like so you can enable So, I mean, right now, it's also already like easy to imagine that you probably don't need those interfaces in a cluttered way anymore, right?
But you want them centralized, right?
Like looking at artifacts and clawed or so.
Do you sometimes feel like, hey, shit, we're missing that?
Or like, I mean, you're obviously like doing your research and you're doing investments there.
But does that sometimes feel that this could be, Also the end of all user interfaces, but one?
I mean, I'm not overly worried.
I don't think we'll get to the world of a single user interface to rule them all.
I think we get to a world of, again, multiple computing services.
Like my Surface, when I'm wearing my earbuds versus I'm looking at my phone versus I'm sitting in my car versus I'm wearing my VR goggles or whatever it might be.
And so in a lot of ways, like I'm already excited because people talk to us about all the time of like taking the Gecko rendering engine and embedding it in a bunch of different places because.
Gecko is fairly lightweight and renders really well.
Can we use that in a car scenario?
Can we use that in an airplane scenario?
Can we use that in a smartphone scenario?
So I think like our technology is sort of like going to all these different places, which gives us a foothold to make sure that all these pieces of technology stay open as well.
So I think that like we all be continuing to proliferate because I don't believe we get to a single place.
I think we get to like every surface is smart.
So then therefore we need to figure out how to enable.
all those surfaces up and down the computing stack in order to pull that off.
I see me chatting with my coffee grinder already.
We'll go back to the days of smart refrigerators.
Yeah, let's see.
I mean, I think we'll see that at a certain point, or at least one device that basically observes your refrigerator.
I mean, there is a world of just like, my phone is an access or a gateway to all these different things.
But I think we have to also assume that not everyone will have a phone.
Or what does it mean in that interim period of just like, what does a world mean of, you know, what does a user interface in sub-Saharan Africa look like versus a interface in Amsterdam look like?
think like in that global view, I think there's a world where we design for San Francisco.
And I think of that mostly like we need to prototype because San Franciscans live in the future, but then we need to figure out what the reality is of the world and make sure we're hitting those two.
Like the beauty again of Firefox is it works on all these different computational, like it works on like this, the cheapest browser on a Raspberry Pi all the way to like my Mac Studio Pro that we're talking on right now.
And so like, I think we need to figure out what it means across everything, not just like the future only.
So do you think that design actually still matters?
I mean, what's the user or what's the ideal user experience and how is that influenced by what we call design now and the looks?
If we now look at, I don't know, my Raspberry Pi.
that just spits out text and is feels like magic because this this text is No, I think we're in this interim period.
I think we somehow got into this place where we think the best possible user interface is a text box that's on my screen.
And that's just crazy, in my opinion.
I think that we just need to...
That's one of the reasons we need this all to be open, because we need more people playing and tinkering.
We need students at design schools.
We need hobbyists in the corner.
We need big companies.
all just tinkering and trying things and putting it out in the market and seeing what sticks.
And I don't think we get that in the world of just like closed AI.
So I want to get to the open so we can get through this rapid experimentation cycle.
Because I think design does matter.
People do want like elegance.
People do want convenience.
People do want like slickness, whatever that means.
But I can tell you right now, we don't have it.
Like none of this stuff in the AI world is elegant or slick by any means.
Like you and I are fine with it because like, I'm also very happy to pull open a Zeesh prompt and just do stuff.
But like my mom isn't, like my mother-in-law isn't.
Like we need to figure out how to make that all work and better.
And I just don't believe that having one company think about it is the answer.
I think having all of us think about it is the answer.
it started off with one, you could say, and now many are thinking about it and now also many individuals are thinking about it.
And I think ultimately, I also see this changing.
I mean, like for Google, for example, it's absolutely logic that you will produce some sort of dynamic applications that make things easier, right?
Like if I want to book travel, then I need an application which is perfectly which perfectly fits my needs, which maybe looks similar than a few other applications out there, but really fits my need of, I don't know, traveling from here to France with my EV car, having the perfect stops, having hotels booked, etc., all of that.
And it's actually also, I would say from Google's perspective, a huge...
uplift and conversion, right?
I mean, looking at from where they come from, from like now clicks, right?
You can stay with clicks even, right?
You don't need to cover the risk of the transaction really.
But having an application that really like goes end to end is kind of...
I mean, I like to think about it as like we can just truly really rethink what it means to serve a person's needs in some ways.
Like in some ways like...
I want to know where my kids are and whether my kids are safe and like whether they're on their bus coming home.
And I don't want to have to go through like five different clicks to figure out the way.
I just want somebody to tell me that they made it on the bus and the bus is going to be here in 10 minutes.
So like, I think like because every surface is about to be computationally intelligent, because all this stuff is being networked in some ways.
there are real opportunities to really re-envision how we get that information to people.
Now, I just want to make sure it's also done in a way that people's privacy is preserved, that our agency is preserved, all that stuff.
But I think we're on the cusp of living in a Star Trek-like world of how these computers interact with us.
And what do you think will that do to us?
I mean, this...
This idea of having a wish machine at home where you can just express, like, hey, I want this, and then it's there, right?
It's really Star Trek-like, right?
What will this ultimately do to us?
I don't know.
So I wrote this piece for The Atlantic a year ago called The Validation Machines, where I was talking about how all these machines, due to the way that they flatter us, due to the way that they say that we're smart and that we're brilliant, eventually ends up with our democracy falling over.
Because it means at some point, you ask this machine, who should I vote for?
And as opposed to you thinking about it critically yourself.
And so I do worry about that world, but I think one of the ways to avoid that world is to decentralize, not to centralize.
And so I think then we can have...
competing incentives.
We can have different people's perspectives come into play, whereas in a centralized world, we're basically all, you know, we're renters, not owners of, like, how that information comes at us.
And so, like, I think a lot about what that tension should look like.
I don't know if anyone knows the answer yet, but we need to figure out ways for this technology to maintain human agency and not take it away from us in some way.
So, and...
Being able to build whatever I want is giving me agency, but telling me I should go vote for someone is not giving me agency, but giving me the information could be.
So I think we need to understand all these different ways to build interfaces, different ways to present information, different ways to be challenged kind of thing.
I want to build a chatbot that just shuts up.
Like right now, chatbots don't shut up because they're measured by engagement.
But imagine a world that they don't have to be measured by engagement, that they're measured by like the right information at the right time.
So they're just quiet 99.9% of the day.
No one could build that right now because none of the business models work, but someone could.
But we need to experiment and play to get there.
Maybe you can build that, right?
With Mozilla.
I mean, that would be cool.
In Germany, there's a kid's book.
I have to think about, like, when we talk about this, it's called Das Zams.
And it's a small animal which is discovered by some guy that has dots in the face.
And the guy just discovers, like, I can wish everything.
It's just happening when that animal is close to me and then one dot goes away with every wish.
And when almost all dots were gone and he had wished really crazy things, then he wished a wish machine.
And I always compare AI to that wish machine now.
And ultimately, yeah, I'm with you that this, I don't know.
I mean, I want to see the positive side as well.
And you just described it.
probably, right?
Like that this will lead to other inventions that maybe don't measure, are measured by the wrong things, right?
And ultimately, I hope for the same, but I sometimes feel like reminded of the, like the time when I traveled to my dad and he asked me like, where did, how did you get here?
And I don't know.
And I remember when we drove to Italy and in my, in my childhood and he had that big book with him with maps inside.
Yeah.
Yeah.
The Atlas.
And, I don't know, for driving, it didn't spoil us really, right?
Apart from us not knowing how we got there, right?
I don't know how it's for general thinking and also for critical thinking, right?
And the risk of really being influenced by… There are opportunities here, and I think we need to enable people to think them through.
Like, you know, I've often wanted to go build… a playlist generator for music and podcasts that looked at my location.
And so like, as I'm like going through the world, it's like telling me interesting things or telling me a podcast about the place I'm driving through.
Like there aren't opportunities here.
The problem is that like it.
defies the business models of the people who are currently building these technologies.
So the beauty of the open web was that people were able to experiment and try, even if the business model didn't make sense at the beginning.
And so I want to get to that world again of just like, we can try all the crazy shit and then something interesting might show up, like a contextual podcast that changes depending on where I'm running in the world.
Like that could be super cool.
And then I would remember where I went.
But isn't that the phase we're in right now?
I mean, we all have that wish machine, right?
And we can, through the subscriptions every day, wish whatever we want, right?
And it's happening.
And won't we discover the new Open Web 2.0 on the way?
Isn't this what will happen?
I think if we go back to what we said before, it's like a lot of this technology forces people to go into their own bubbles.
Of like, I built the perfect thing for me.
But what we need to figure out is how do we still reconnect together as a society and reconnect together as people?
And so none of this technology is enabling us to talk with each other.
So I'm building all my own apps.
You're building all your own apps.
But when do we share them?
When do we talk with each other?
How does that work again?
And so I think we are getting close, but we haven't broken through.
We all have this strong desire to show off.
what we've done in the world.
There's a reason why we do show and tell when we're young children in school.
And that's the reason why open source software works so well for a while, is people just want to show off and then collaborate.
These technologies have a tendency to cause us to go into our bubble.
We need to figure that out.
That's one of the hard things that I think, like if you were looking for something that you could be pessimistic about, it's just like, if we don't figure that out, we could have a problem that we all live in our own bubbles after all this.
But isn't that, in a way, also something that the big labs have to think about?
I mean, they all have that problem as well.
But why do the big labs care about it?
Like, the big labs only care that you, Toby, are using my technology.
No, they want to suck you in as a company.
I mean, I think that's the big topic, right?
I mean, that's Entrofix's mission, I would say, that they want to make the whole...
company and they have to make the whole company more productive.
It doesn't help if some random guy produces some software that no one uses.
They have to get to adoption of the software as well.
They have to also do something for distribution.
Otherwise, at a certain point, people just say, hey, I can use the Chinese stuff, right?
Or I can use whatever.
And I think this makes us ultimately aligned in goals in a way.
And I would say it's kind of the next step, right?
Like right now, I don't know, I can produce the perfect to-do list for myself, but I have to produce the perfect to-do list and the payroll tool for my whole company.
Otherwise, it doesn't work long term.
I understand what you're saying.
And like, I guess like that doesn't help us though on like the societal aspect of it.
I think it helps us like inside a single company.
But like, I'm worried about then how do we then like move together as like a group of folks, not just like each.
companies are like figuring it out for themselves.
I think maybe it has to be in a way that like people just reconnect in real life at a certain point.
Yeah, maybe.
And have those moments, hey, look at what I've built.
Hey, look at what I've built and challenge each other.
But yeah, I'm with you that this is kind of worrying and really something to solve.
And what would be cool if you can solve this with Mozilla maybe?
Yeah, I mean.
It would be an honor to try.
Let's try it.
So coming maybe to a few advices that you would give to other CTOs and ideas that you carry on before we come to the end.
So two years out from your perspective, which roles in engineering look really different?
And which ones do you think look exactly the same?
Is that none?
All?
I mean, I think they all change.
I think our complete definition of just what software engineer, product manager, product designer, SRE looks like, I think all gets completely upended.
So it's really hard to predict, I think, what we need to do in two years.
I think we need to look at it more like an incremental level.
And so I think the thing I would tell CTOs today...
is to start really thinking about both what does AI penetration look like on your teams right now, but also like, what's your backup plan?
Like, what do you do in the case of like, how do you make sure you're not wedded to a single provider?
How do you have multiple buyers?
How do you bring in multiple perspectives and multiple tools in so you're actually evaluating it all yourself?
Like, you know, I used to build data centers as part of my job at Twitter, and you would never allow you to just have one power line into a data center because like...
a squirrel to that line, it's going to be a bad day for the squirrel in the data center.
But like, you need to have two, you need to have two lines coming in, which then gives you choice.
It then gives you leverage.
Same thing happened when we had cloud, right?
Like everyone built AWS and then AWS jacked this bills and then it's hard to leave.
So now everyone uses like Terraform or something that allows them to have abstraction that can move my code, which gives you negotiation leverage.
So I think for CTOs today, Like literally right now, I would think about what gives you leverage in this world, not just leverage in the sense that like it makes your team more productive, but leverage in the sense that like when your team is more productive and they're addicted to these tools, how do you make sure that you can move?
How do you sure that you can change?
And then I think having that kind of leverage allows you to have more control over what happens over those next two years, as opposed to just taking like the anthropic recipe of what your team should look like and just.
having to run their playbook because you're locked into that tool system.
But isn't it in a way also what happens right now, the technology, isn't it too simple so that you can almost always switch?
And isn't that in a way also the big problem of the labs?
I mean, but it's not super simple, but it's not.
I mean, so memory is one.
It's the big example, right?
Like if I use Claude for a week and then switch to ChatGPT, it's like talking to an amnesiac because.
Claude knows everything and ChatGPT doesn't.
So already right there, they're creating some form of lock-in.
And then right now, Claude has their ability to share a session with another user so that they can then take over your Claude session and work there.
So they're already trying to lock you in by having you upload all the data to their system.
So I'm saying that, like, could you...
Could we have abstraction layers that means that memory is on my side or you're reading data off of my file system, not off of your file system?
Like, how can that all work so that then choice remains on the side of the CTO?
Yeah, I think this is obviously true.
Like, I mean, I try to avoid all of this, right?
I try to avoid using memory as good as possible, like using Claude's memory, not my own.
and try to really stick to markdown files, etc.
I guess you, like, what do you do then?
Like, the same or how neat does it go?
Well, I mean, the benefit of running local models is that, like, there is no place for that memory to go.
So that memory is just being written either as Markdown or HTML onto my local file system.
But it's the question of the big frontier labs.
Like, they're incentivized for you to, and they can provide you better products.
I'm not faulting them, but they're incentivized for that memory to be sucked in.
But so, yeah, for me, one of the projects that we're working at Mozilla is a portable memory standard of what it would look like to actually have memory and then have all the different characteristics that you want.
Like, how do you have the ability to turn on or off different memories?
Do you have the ability to forget something?
Like, what does that really look like?
So like one of those are one of the collaborations that we're starting to do across the industry.
how should we build systems like that, which then enable us to then move between providers relatively easily?
Okay, okay.
Yeah, I mean, that absolutely makes sense.
And there are also like a lot of more experiments on the long-term memory side, right?
Like, for example, G-Brain, I don't know if you know that, like all of that is also like helpful.
And I think also really space to have something bigger as well that...
really works, right?
And that really brings value.
I'm not convinced to be all end all are Markdown files.
Like, I think like they're the good kludge for right now.
But I think like, you know, is it the world that we're all training our own model?
That's our memory.
I don't know what that architecture looks like, but it just seems like this is a, this is like a, we haven't even gotten to the database world.
We're still in the plain text world, like kind of thing.
And so like, I feel like there's still a bunch of iterations that need to happen on the memory side.
Okay.
And what else?
Like memory, I guess, then open code, right?
Like you would recommend as a standard, what else?
What standards would you, or what standards do you want to set?
Yeah, I mean, I think there's still questions about what do gateways look like?
I think there's still questions of what do RAG systems look like for storage?
I think there's...
questions about what tool calling, execution environments, all the stuff that we've sort of taken for granted because the closed systems have provided it all, we need to sort of recreate that in that open ecosystem.
Thankfully, lots of people are doing it.
This is not just a Mozilla problem.
Lots of people are working on what open tool calling, what open sandboxing, all that stuff looks like.
What needs to exist is just a little bit more glue to make it really easy for you to just like...
app get install something and have it all show up on your servers.
I think the good side is, or the good thing this time is also that there are like a few standards that are actually being created on the way, right?
I mean, we could debate now, it's like MCP versus CLI, like what are the benefits?
But also, I mean, there are at least some ideas and standards already created that are like...
everyone can use.
Yeah, so I think tool calling is one.
I think execution environments is one.
I think MCP is going to be another one.
I think the interface between the model and everything else is mostly the OpenStreams guys, but that is one of them.
The storage layer is another.
And so I think we can start to create what a LAMP stack or an OSI stack looks like, and each one of those layers is going to need, you know.
I'm not naive in saying that we need to do what the web did, that we're going to have a whole bunch of people with white hair on their beards arguing about standard.
I think we will build, and then from there we will abstract out what that interface looks like.
That's what happened with OpenStreams, right?
We basically took the chat GPT and the...
the OpenAI model interface and standardize that and then start adding extensions for the things that it couldn't do yet.
So I think like as we're starting to build all these pieces of glue, the standards will start to emerge because people are going to want to be like, I like how these two systems are plugged together, but I want my own bespoke system for this one.
So I'm just going to swap it out and then that standard will coalesce.
Yeah, looking forward to have more open standards around it.
So you just said one sentence connected to, I think, renters, not owners.
And you're writing a book, Owners, Not Renters.
Can you tell me what this will be about?
It's not finished yet, right?
It's not finished yet.
So, I mean, what the book is sort of exploring is what does it mean to have technology that's on your side versus someone else's side?
Social media is one example.
I think what the AI system is another example.
But this happens in the physical world all the time, right?
So like John Deere tractors in the US are a good famous example.
Like people are buying John Deere tractors from the 1970s right now because they can fix them.
So like I paid a couple hundred thousand dollars and I can fix them.
But if I buy a John Deere tractor from yesterday, I can't fix it.
If something goes wrong, I need to bring it to someone because they don't publish the specifications.
There's no way to open it under the hood.
It's the whole like, is the whole fixed revolution that sort of like broke down somewhere.
So what does it mean to live in a world and who are the people that are building that world of technology that you can truly own?
We're seeing this on streaming media, right?
Like you go to Netflix and then one day all my favorite TV shows are gone because the licenses disappeared.
And there's this whole push to get to own the media again of people actually starting to like...
create their own DVD libraries again because they don't want that stuff taken away from them.
And so like, how do we get back to that world and how do we get back to the right balance of things being owned versus rented?
But I think that AI is actually going to help us there, right?
I mean, look at my coffee grinder example.
I mean, my coffee grinder could be like just another device that is bricked by some manufacturer because they have been acquired and now they don't operate the platform anymore or whatever, right?
And I can, I don't know, whatever has an ESP32 chip, I can own, right?
Sure.
Or it could be that the model that you use refuses to do it because they've determined it's against their terms of service to do that kind of decompilation and recompilation of the firmware.
I don't know.
And so that's why I like to open models because I can just download them.
They're under my control.
Yeah.
Yeah, I mean, as soon as that happens, that, I don't know, Claude denies to, I don't know, decompile this or that, I would also switch, right?
I mean, that's common knowledge now.
And I think that's going to be quite exciting that a lot of modern things can also be owned again because the world is so complex now that you can dramatically simplify it through AI.
I think I'm not saying that the world we live in is bad or good.
I think we're in a change of massive, we're in a time of massive flux.
And so like we need to figure out how to resettle down to whatever values we want to be in the world.
And I want to be in a world where we own things, not rent things, like you said.
I don't know the exact path to get there, but I think the first thing we need to do is just agree that we want to get to owning, not renting.
And then once we get there, then we can start to plot the path to pick it, to pull it off.
Let's agree to that fist bump.
So little surprise for you.
Somebody on the Firefox team told me there's a hidden page almost nobody outside of Mozilla knows about.
And you basically have to type about rewind into your URL bar and you get a date picker styled like the original Phoenix UI.
And you type a year and it takes you there.
We are now sending you back to 2009 when you were starting as a VP of platform at Twitter.
And you now have the chance to whisper something into young Rafi's ears.
What would it be?
Yeah, I would talk to him a lot about what it means to have open platforms.
Like in 2009, we were opening up the entire Twitter API to the world.
And then what happened two years later is we basically shut it all down again.
And so like we took away access to the Twitter platform because it got in the way of advertising.
So like it meant that people could build clients and not see ads.
And so we actually had to shut it down again.
What I would have told Rafi in 2009 was to not let that happen.
Actually, what we did was reduced innovation in the ecosystem, and we traded that for short-term profit on advertising.
What instead we should have done is kept that open, as open as possible, and even more open than we were doing, and figured out a way to monetize.
There are so many different business models that at that point, we were locked into trying to do the Google business model, but on Twitter.
And we needed to not...
get stuck in that mindset.
We needed to be creative and embrace what openness could have done.
Like the goodwill through a developer ecosystem, the opportunities to integrate in 1,200 different places were all taken away because we were trying to follow the one model of Google.
And so that's the thing I would have talked to Rafi about, just like, what does it mean to shut down an open API and what it could have been instead if we just kept that open for the world?
And, I mean, you're always smarter if you look through the rear mirror, right?
It would have kind of probably saved Twitter as well, if you look at the history, right?
Yeah, I don't have enough hubris.
I mean, I built one of the versions of the Twitter API.
So saying that would mean that my work saved Twitter.
I'm not going to say that.
But I think that like...
A lot more interesting opportunities would have shown up that will allow Twitter to expand in a bunch of different ways that it was just constrained to do because we were following the Google Playbook.
Makes sense.
Makes sense.
So don't follow the big techs, do your own thing.
That's the key message today.
Rafi, thanks a lot for being on my podcast here.
I'm really looking forward to publish this.
Been a lot of fun.
And let's see what follows the coffee grinder or...
Yeah, check with me in a year and we should talk about the coffee grinder.
Let's do that.
Have a great day, Rafi.
You as well.
Bye.
Thank you for listening to the Alphalist podcast.
If you liked this episode, share it with friends.
I'm sure they love it too.
Make sure to subscribe so you can hear deep insights into technical leadership and technology trends as they become available.
Also, please tell us if there is a topic you would like to hear more about or a technical leader whose brain you would like us to pick.
Alphalist is all about helping CTOs getting access to the insights they need to make the best decisions for their company.
Please send us suggestions to cto at alphalist.com.
Send me a message on LinkedIn or Twitter.
After all, the more knowledge we bring to CTOs, the more growth we see in tech.
Or as we say on Alphalist, accumulated knowledge to accelerate growth.
See you in the next episode.