Brave's Story with Jonathan Sampson

So, this is pretty exciting. Eat Watermelon is my first one. I think we have some people who have This is their second time around or so. Um my name is Sampson. I work with Brave.

So, today I'm going to briefly tell you the Brave story, what it is, how we developed. I think there's already quite a few people that are familiar with Brave that are in this room. We're just having a nice friendly chat beforehand. Um so, quickly, I guess a little bit about me. Uh I started programming like 23 years ago, which is getting like more frightening each year.

And so, I started with this game, Jedi Knight: Dark Forces II. It was really cool because in the '90s, uh software was not as complicated. It was It was actually quite a bit more open. It wasn't open source necessarily all the time, but there are components of it that you could view. And so, this game had these things that were called cog files.

They were written in this cog scripting language, and you could see object references and stuff in there. You'd see like door.3do, and you could see events like open or, you know, player initiate. And so, I was not a programmer at the time, but I was able to read these English plain text files and start out patterns, things like door.3do.

And there was a character in the game called Max. There was a max.3do, which is a little rabbit. And so, I realized if you change the reference from door 3DO to max 3DO, save the file, refresh the game, whenever you like shoot your uh No, no, it was a bullet. If you change the bullet reference to the max 3DO, you'd shoot your gun, a rabbit would pop out the end.

It was this weird stuff that you could do, but I loved it because it was so simple. It was easy enough that anyone could hack on if they wanted to. And so, this is kind of the entire way that the web itself in the early '90s was. This is NCSA Mosaic from the National Center for Supercomputing Applications at the University of Illinois Urbana-Champaign. This was probably the the most popular browser um to ever exist.

I mean, when you look think about global market share-wise, uh this was the work of people like Marc Andreessen and his uh fellow teammates. He was making minimum wage at the University of Illinois Urbana-Champaign uh when I think it was Jim Clark came over and pitched the idea of starting Netscape Navigator and then overnight Marc Andreessen was worth $600 million. So, that's a nice little pay bump right there. But, I'm curious, has anyone in this room, I don't know people's ages and stuff, has anyone ever used like a browser in the early '90s or mid-'90s? I feel like one person, two people.

So, I don't know how many of you have ever got to enjoy like this wonderful serenade. This is that that dial-up tone. What I love about this though, I never knew what this was beforehand. This was just some like crazy ritualistic music that played like some early '90s uh, like EDM or something. And uh, it's actually the negotiation between the device and the server over what protocols are supported.

I'm going to stop that right now. And you can actually hear the devices negotiating what protocols are supported and then figure out how they can communicate going forward and pray to God no one picks up the phone receiver downstairs and disconnects you. And so, that was around 1992. And the next few years the web got really exciting because of some unrelated innovations from a few people that all came from the University of Illinois Urbana-Champaign, coincidentally. In 1993, Marc Andreessen, who we talked about, he was again a student there working in the NCSA.

He wanted to put images inline on web pages. So, as we saw with that first web browser in the Mosaic, you had text, you had hyperlinks, you could click those, you could open up a second window and you could show a picture in the window. But, they wanted to put the pictures in the pages themselves. And so, what Marc did is he actually went ahead and just did this work and then after it was basically done, he pitched the idea to a a forum or a talk group and said, "Hey, we should do this." I mean, he already did it, but he was trying to I guess somewhat be fairly formal.

But, the way he did it was that images could be loaded from any third-party server. So, if I had my server with images on it and you wanted to show the images on your website, you could just load them directly from it. You didn't have to have this file redundancy, where you're multi- you're duplicating these files all across the servers. It's really cool. They developed this this term called hot linking because the idea was if too many people tried to pull images or files from your server, it would heat it up and melt it.

Um but it was really neat that they avoided that redundancy at that time. The very next year, Lou Montulli, who is a colleague of Marc Andreessen's at Netscape Navigator, Lou was working with a guy named Vint Cerf. Vint Cerf is the pioneer of TCP/IP. Vint is unbelievable, and he's still working to this day at Google, I believe. And so what Vint did is he he told Lou, he's like, "We need to find some way to put just a little bit of data on the user's device, send that data out with every request, allow the receiving servers to modify that data, and when it comes back, store it on the user's device again."

This would give us the ability to log in to websites, to persist sessions over browser you know, sessions. So if you close your browser, open it back up, you don't have to re-log into a site. Or you could remember what's in your shopping cart, these types of things. Completely unrelated to images, Lou did this work. He was inspired by the HTTP or actually it was he was inspired by the magic cookie on Unix boxes.

So he created the HTTP cookie. The the patent He filed a patent for this. The patent is so cool. It's an arrow pointing to the right, and then it says like server modifies data, and then an arrow pointing to the left, and it's like machine receives modified cookie. And that's the patent itself.

It's It's the craziest thing in the world. And then 1995, this is when the world was set on fire by young hacker named Brendan Eich. This is our CEO. He would go on to co-found Mozilla as well, and create a Brave. And so Brendan was tasked with creating a like a baby brother scripting language for Java, hence the name JavaScript.

And of course, this just really changed things for the rest of history. 1995 was an exciting year. In 1995, a company by the name of DoubleClick was formed. Um later on, a company by the name of NetGravity was formed. I think actually NetGravity might have come before DoubleClick.

It's tough to tell. Uh but what they started doing is they saw the potential between what uh Marc Andreessen had done and what Lou Montulli had done. And they realized that if I had a popular image on my server and everyone was loading that image from my server, then everyone was sending cookies to me and I can start to monitor people as they move between sites that load my images. And so, in 1995, DoubleClick uh was created. NetGravity was created.

This was the work of John Danner and Tom Shields. And they had boasted that ad server can dynamically target ads based on the following criteria: browser type, computer platform, the domain, the time of day, location, proxy server, search terms, keywords, cookie certificates, more information, and even user registration databases. So, they realized there's a lot of meta information that comes across with every request. So, when your computer tries to load a picture, it sends an HTTP request. It tells some information about you along with a cookie.

When it tries to load a script file, when it tries to load a page, all that stuff contains meta information that is necessary for making the web work. Um this was their website back in like 1996. It's beautiful as well. And then, so in 1996, we also started seeing journalists and tech-minded people realize the the potential for harm here. And so, this was published here on Thursday, June 13th, 1996.

Um this paragraph in particular, DoubleClick boasts for about the precision of its methods. For example, the company says, "With DoubleClick, you can target females who work in the aerospace component manufacturers uh in the Los Angeles area with at least $100 million in operating revenue who use PCs, Windows 95, and like sports." So, in 1995, um 1996, this was incredibly granular and this was still just getting started. So, we also had heroes back at that time. The Netscape Navigator browser, they introduced the Netscape plug-in API 1995 as well and people quickly co-opted this technology to start to build some type of defense against this this monster that was developing that was just harvesting people's information.

So, these fellows right here led by James Howard, they created a company called PrivNet and a product called the Internet Fast Forward and the whole goal was to identify images based on their dimensions. So, if it's you know, 720 by 60 or whatever common ad sizes were, they could stop those from loading. They could prevent cookies from loading. They'd give the users the ability to filter. This all was happening in 1996, but their motivations for this their motivations were quite simple.

They said we just didn't like the time it took to download advertisements while we were paying for our time online. I don't know if many people have ever paid for their time online. They pay for access, but when I was online in the 90s, I got 15 minutes a day because we had to pay for the minute. And so, I got 15 minutes a day on the internet. So, I would have been with these guys right here.

I definitely don't want to spend that time you know, loading advertisements and stuff like that. We'll come back to this over here too. This was kind of coincidentally they had this uh uh projection here for what advertising in cyberspace would be in the future. And so, we'll come back to that. It's pretty interesting.

So, we talk about the you know, time is money type of aspect. Uh in 2015, the New York Times did a research a little bit of study on the top 50 websites that Americans visit and how much of that data goes towards content they actually want to consume versus how much of that data goes to ads and trackers and other tertiary scripts that are set up just to monitor and surveil them. Um and the vast majority you can see with Examiner, you know, you're paying almost 10 cents for trackers and only like 2 cents for actual content. Some of the really egregious cases here, more than 30 cents of your data is spent loading trackers and ads on boston.com as opposed to like 8 cents to load the actual content.

And so there's massive massive waste in this space. In fact, I think they concluded that Americans like myself would spend on average $23 a month to be tracked. And so that's not just for the access to the content. I'm paying to be tracked. So, what are we getting for this?

Um, we go to a web page here and there's an ad slot, this empty region right here, and once that ad slot is parsed and processed by the the rendering engine and so forth, um, it sends out this communication to ad exchanges who then in turn communicate with demand-side platforms who then in turn communicate with data management platforms. This entire process is emanating private data about me out into this sea of faceless third parties. Information like my gender, my income, where I live, religious affiliations, marital status, all this type of stuff. All that information gets sent out and an auction has taken place where I am being put out in front of a bunch of advertisers to see who will pay the most for me. And of course this all happens within 100 milliseconds.

And once they find the highest bidder, an ad is shown that they think will target me and I mean it might. And so that's what we get for this entire exchange. That information is not temporal either. It is persistent by these third parties. They can hold on to it, build up little dossiers, and build up these profiles moving forward.

Facebook did this many years ago that I'm assuming they still do where they build up shadow profiles. So they'll collect information about people before they ever become a member on Facebook. And then once they become a member, they can adjure them to all that other data they've already been collecting over the years. So, in December 2018, the New Economics Foundation had said that bid requests on UK users are being sent at a rate of almost one 10 billion per day, or 164 per person per day, and are seen by hundreds if not thousands of advertisers. And again, it includes information like my health.

I might be looking up a rash or something like that online. Religion, my age, all this type of stuff. 10 billion per day. That was in the UK alone. 2017, um we started realizing that we're not just getting ads, either.

And these aren't just pictures that are annoying. It's not that like jiggly belly fat animation or something or some type of weird parasite in a fish's mouth. Like, these are really harmful sometimes, but they didn't always manifest themselves as being harmful. So, we saw with CoinHive in 2017 that people started finding out ways to get these miners JavaScript-based miners into iframes on web pages. And so, the Showtime websites were secretly mining using user CPU for cryptocurrency.

That was from The Verge. October 2017, hackers have turned PolitiFact, a political website, into a trap for your PC to mine cryptocurrencies, Monero in particular. And then they adjusted their threat a little bit, and they started going after Chrome extensions. And so, Chrome extensions were being hacked to secretly mine cryptocurrencies because obviously, it's it's a a whole 'nother area for these people to spread. In fact, we saw last year, I think it was, MetaMask had an impostor extension in the Chrome Web Store.

And I was talking to Dan Finlay about this. I was like, "Hey, there's a fake MetaMask in the Chrome Web Store." And so, Dan reached out. I believe him and his team reached out to Google. Google deleted the actual MetaMask and left the fake one in there.

And so, it is unbelievable how risky even extensions can be these days. And of Of there are many extensions that people use that are no longer maintained that if I were to come up to the owner of the extension say, "Hey, I'll give you $1,000 and over the the control of the repo control of the extension itself, they'd be more than happy to do so." And then of course, the next time I ship an update, that update goes out to all of the existing users and you have something like Coin Bit you know, Coin Hive in your browser. Government websites as well, 2018 were hit by cryptocurrency mining malware. Um several other cross-site scripting attacks and stuff as well.

Uh January 2018, this was the creme de la creme. This was these hackers were able to get it into YouTube ads. So, we think that it's only the shady little advertising networks out there on the the the the external you know, the outside of things, but YouTube itself were compromised uh by by these cryptojackers. So, how does the how people responded? By and large, we've responded by blocking things.

More than 600 million devices had ad and tracker blocking on them back in 2016. Um mobile started around 2015 and of course, it is just completely eclipsed desktop. It continues to be that way today. But if we're blocking all of that stuff, let's see. Actually no, this is really interesting.

So, in response to the the increased blocking from users, these advertisers and companies that profit off of this capitalistic surveillance, um they've had to maneuver themselves in different ways to continue making money. So, Google and MasterCard cut a secret ad deal to track retail sales. This is insane. This is what they had said is Google had access to approximately 70% of US credit and debit cards through partners. 70% of credit and debit cards for people like me, Americans that are down there.

Since 2014, Google has flagged for advertisers when someone who clicked an ad visited a physical store using a location history feature in Google Maps. So, now it's not even in the browser. It's like this phone that I'm carrying around that has Google Maps on it is tracking my location actively. And when it sees that I'm near a location that was affiliated with an ad that was displayed to me, Google is able to correlate those two events and then inform the advertiser that they need to pay a little bit more because Google is so awesome or something. Um and of course it's they're not stopping there either.

Uh they said this should be a third quote. Here we go. Google has continued to push deeper into offline measurements. The company like Facebook and Twitter Incorporated has explored the use of beacons or Bluetooth devices that track when shoppers enter stores. So, the actual devices are tracking our location as we move about physically from store to store.

In fact, I mean you can see this the other about a month ago or so, my wife and I went to the bank to talk about to our banker about some issue. And uh we spent about 30 minutes there. We stepped outside. I opened up Twitter and the first ad I got was for a loan from a competing bank because geographically my latitude and longitude said I was in a bank for 30 minutes. So, that means I'm probably looking to get a loan or do something bank related.

And of course, when that was uh sent out to these ad parties these uh third parties, they decided that I would be a a good candidate to show an ad to. So, as a result, Brendan Eich, the guy that created JavaScript back in 1995 1994, um he got this idea for Brave. He figured something needs to be done. We have to figure out some way to make a private yet sustainable internet because we again have 600 million devices blocking ads and trackers. That's wonderful from a security perspective, but it doesn't do much for the sustainability perspective because we still have these websites that are losing massive amounts of revenue.

And so, Brendan had called on his buddy uh Brian Bondy who was also at the Mozilla back in the day was at Khan Academy whenever they got the the idea to do Brave together. And on January 20th, 2016, we had our inaugural tweets, "Join Brave and make it our internet." This is when the excitement really began. So, what is Brave? Brave is a privacy-oriented web browser that focuses, again, not on just user security and privacy, but on the sustainability of the web.

So, as you browse the web, we have baked into it this logic. This panel actually here is called the Brave Shields panel, but it actually blocks third-party trackers by default. It looks to preemptively upgrade insecure connections to use HTTPS when possible. It blocks third-party cookies, and it blocks third-party device recognition, that type of stuff. They'll use the Canvas API to figure out some type of way to generate a unique token for your device for use so that they can monitor you as you browse the web without having to use cookies or something like that.

So, Brave blocks all that right out of the box. Jump to the next one. Okay, so when we're blocking this stuff, we have to think again about the the revenue aspect of this. So, this comes into that sustainability model. You can't just block for a better web.

You have to also think about how do we take care of the people who are writing the Solidity, you know, tutorials on Medium, or people who are contributing to open source and that type of stuff. How do we take care of these individuals? And so, what Brendan had the idea was to create the Basic Attention Token. He had visited the Ethereum offices in London, I think in 2015. He already had the idea for some type of Brave coin at that point.

And then realized that, you know what, the ERC-20 token's probably the route to go. And of course, we initially launched using a a a prototype built on Bitcoin. I think we were talking about this a little bit before we started the broadcast. Um and that was working fairly well for a while until about 2017 hit. And then Bitcoin congestion and Bitcoin fees became insurmountable, and people who wanted to buy $5 in Bitcoin might have to pay like $7.

50 in fees to do so. And so, that simply didn't work. So, we jumped the laterally over to the um over to the Ethereum blockchain, which has been fantastic. So, now as you browse the web, you come across a verified publisher like the Washington Post, you can expand your rewards panel. I can see they're Brave verified, so I can send them a tip if I want to.

I can choose between 1, 5, and 10 BAT. These are the default values. The publishers have some liberty. They can also make it 5, 25, and 50, I think, or something like that. Um and I'm able to send that using the Ethereum blockchain, which is pretty magical.

Uh but we realized that it's not just traditional markets. It's not just journalists that are producing some of the information that we find valuable on the web as well. Sometimes it's someone on Twitter, it's someone on Reddit, it's someone on GitHub who just has a really awesome commit. So, Brave is actually looking to create sustainability across a an array of different platforms, including Twitter, GitHub, Reddit, as we mentioned before. So, we come back to these guys.

Um the advertising in cyberspace. I thought this was interesting. It's kind of just a side note. Uh but we can see what they projected at the time and what we actually saw in retrospect. There we go.

So, they got it right for the most part, but then the actual, you know, explosion happened there uh between '98 and '99 and uh and things really start to be they are over performed over performed far beyond the expectations of people in the industry at the time. Today, I think we're spending We're seeing more than a hundred billion dollars in revenue digital advertising and like 20 billion of that uh is going to fraud each year. It's unbelievable. So, what do we see? This is the comparison between the incumbent digital advertising system and what Brave is suggesting.

Modern digital advertising, you are not asked to participate. Your data is captured, recorded, shared, and sold multiple times per page load each within a hundred milliseconds. Uh you're auctioned off before a panel of faceless third parties. The highest bidder gets to show you an ad. Middlemen and fraud capture most of that revenue.

What Brendan has uh conjured up here, and which has been working extremely well so far, uh is the Bri- the Brave's private ads alternative. You decide whether you will participate, to what degree. You can say, "I want to see up to five ads per hour, or as little as one ad per hour." Your data never leaves your device. This is what I love about it.

He has uh he's come up with a model where you completely invert the the incumbent system. So, previously, what we have is your data gets sent out to this sea of third-parties, as we saw earlier. All of that collection allows them to make very fine-targeted advertisements, uh but at the cost of you losing control of your data. If you invert that, and all of your data stays on your device, and instead all of those creatives from those advertisers are sent to your device, your device can leverage local machine learning to figure out which of these ads, if any, is relevant to you, and your data never leaves your device. So, you get better targeting without that leakiness of the uh incumbent system.

So, we use machine learning, and of course, if any ads are shown to you, if you ever see a notification, 70% of the revenue, so 70% of what was, you know, gone in to show that ad to you, goes directly into your wallet. That's the magical bit right there. And of course, we had, I think it's like 14% conversion or click-through rates was the most recent number, which is seven times more than the industry average. Sure. Um what So, what doesn't go to the user goes to the other parties involved.

So, it's actually the lion's share. 70% goes to the ad slot owner, and 70% is pretty close to being like an industry average. If you look at the um iTunes store and that type of stuff, it's all around 70% that goes to the actual publisher or the creator. And so in this case, we figured, "Okay, we'll follow that. We'll do 70% to the ad slot owner, which in the case of desktop private ads is you, the user."

In the case of an emerging ad network which uses the or an emerging model which uses publisher pages, that type of stuff, where we'll actually show ads on pages, the 70% would go to the publisher in that case, and then you, the user, would get 15%, Brave would get the final 15%. So, I think the 70% there was precedent already in the industry for that being a pretty solid number. If you come in anywhere below that, um I don't know that you're being very competitive. Um so, I hope that kind of answers the question. Thank you.

I can We can talk all day about this. Sure, yeah. So, 70% of that goes into your wallet inside the Brave browser. Again, this is using the Basic Attention Token, the ERC-20 token. Um you can then take that out if you want to uh through uphold.

com. You can convert it into some type of fiat. You can exchange it for a Bitcoin or Ethereum or something else if you want to. It's yours. It's just a token on the network.

Um and so on the desktop, if I'm browsing the web, I'm looking at the ETH Waterloo website, and in the background, of course, I'm trying to figure out like You can't really tell, but I'm looking for ways to convert ERC-20 tokens into burritos. Uh and so if I'm doing that, I might get an ad. The browser itself churning on that ad catalog that it downloads, the machine learning module will actually figure out, "Okay, this guy is looking for ways to use utility tokens, blockchain assets in such a way that he can purchase reward goods, burritos in particular." And so the Ternio BlockCard is advertising, and they let me know that I can actually get a debit card or credit card type of thing that lets me spend my utility tokens anywhere Visa is accepted. So, if I click on that, I get a full-page pop-up.

So, they get the entire window. And this is actually really interesting me if you look back into like 1994, we didn't cover this earlier, but 1994 is when pop-up ads started to exist. And there are some people that are young enough today, they've probably never seen a pop-up ad, but they became a rather big nuisance, but they were created for a good reason. There was a um there was a company, I think it was Tripod actually at the time, tripod.com, but there were some websites.

There was a car dealer website that started having adult ads placed on his page. And he was like, "That's not cool. That's contrary to our business model. I don't want that here." And so uh what they did is they uh one of the engineers at Tripod realized that we could just pop the ad out in its own window and you can't associate it to any other content.

And so you have brand safety. And of course, pop-ups became co-opted. Um they were really bad at that time. You could use JavaScript to move the window around the page. And so you could make some really annoying pop-up ads.

Um and so uh we got rid of pop-ups. Ads went back into pages. And of course, we see the value in having this brand safety. In fact, a couple years back, we were at um uh was it Mobile World Congress. And one of the main issues, I think it was actually a year ago, one of the main issues advertisers there were looking for was brand safety because it's it's become increasingly difficult to deliver on the web today.

So Brave does that by just giving the advertiser their entire window. They get the entire tab themselves. They can do whatever they want. With the exception of loading third-party trackers and that type of stuff. There's no special privileges within the the Brave system.

And so today we have over 9 million monthly active users. Just a few of the properties that have verified publishers on 16,000 Twitch streamers, 32,000 Twitter users, 35,000 websites. The YouTube number continues to blow my mind. Nearly a quarter of a million YouTube channels are verified to receive receive Basic Attention Token. And of course, Reddit is fun.

If you want to upvote someone for their awesome memes, you can give them BAT now as well. And that's pretty much it. There is a few things that love to ask people to do if they're watching or if they're here in the room, download Brave, opt into the Brave Rewards program, participate in Brave Ads, contribute if you're an engineer through github.com/brave. If you're a hacker, which there are quite a few at this event today, if you find a vulnerability or something that just isn't right in Brave, you can submit it to hackerone.

com/brave and we will likely pay you for that. Um and that's basically it. A few URLs you can check out, brave.com/download, brave.com/tips and if you have any questions for me, twitter.

com/bravesamson. So, thank you all very, very much.