Presearch on the CryptoCanucks Podcast: Fundamental Shifts & Protecting Privacy

Presearch founder and project lead, Colin Pape along with David Howitt the team lead for business development, sat down with Kohji of CryptoCanucks one day to talk all things Presearch, the state of crypto in Canada and the importance of protecting privacy.

CryptoCanucks Podcast is the leading podcast about Canadian cryptocurrency & blockchain startups. The podcast aims to educate people about the various blockchain startups within the Canadian community, as well as, the basics of blockchain technology.

What follows is a loose transcription of their conversation. You can catch the entire audio at the end of this article.

Kohji for CryptoCanucks: “Episode 3 is with some guys from Presearch.io, David and Colin. What is Presearch you might ask? Well, they’re going to explain it, but it is an open, decentralized search engine that rewards community members with Presearch Tokens for usage, contribution and promotion of the platform. We go into a little bit more detail in the interviews, so check it out. And of course, let me know what you think…”

Kohji: “Welcome to another episode of our crypto-show. I am here with the, is it fair to say, ‘creators’ of Presearch? They are nodding their heads for the listener at home. If you guys want to introduce yourselves, so people can get used to your voice.”

Pape: “Sure, my name is Colin Pape. I am the founder and project lead at Presearch.”

Howitt: “And I’m David Howitt, I’m the team lead for business development.”

Kohji: “Awesome, OK. I think the best place to start is probably an introduction into what you guys do, cause for the most part, we’ve so far been talking to people that are mainly currency based, but I think your project is definitely more than that and I’m not going to do it justice if I try to explain it so I’ll let you do it for me.”

Pape: “Yes, so the concept for Presearch is basically a decentralized search engine, search platform that is built by the community for the community. We kind of refer to it as the ‘People Search Engine’ and something that has the transparency and leverages open source technology that is kind of lacking within the current search eco-system.”

Kohji: “Ok, I think the first obvious question, and to a lot of tech people they are probably just going to say why would you even ask this, but why do we need a decentralized search? I mean I search for, you know, just before we started my microphone was broken and I was like ok, ‘new microphone’ in Google . New microphone popped up, I clicked on it, I bought it. So, why do we want a decentralized search engine?”

“…it’s really more about decentralization from a control and power standpoint.”

Pape: “Yeah, so I think there’s really, technological decentralization, and that’s part of what we are doing, but it’s really more about decentralization from a control and power standpoint. And if you think about it, all of the systems within the world, including all the governments, the most centralized system is search. And it is the most influential at the same time, because people do rely on it to find things that they are looking for, they put a lot of trust into it and they rely on the fact that it’s going to provide the best results and there are kind of a lot of different definitions of ‘what it the best.’ Currently Google has 96% of the search market space so it’s extremely concentrated. If you’re using it strictly as a consumer, you may not notice, but when you are a producer you are operating within the ecosystem as a creator or somebody who is looking to connect with potential users or consumers that is really where you run in to this massive centralization and it has the ability to make and break different services, prevent people from entering into different services and it really limits the competitive landscape. So we’re taking a multi-prong approach to it, but our experience was kind of born from being marginalized within Google and seeing the need for an alternative. And something that is more open and transparent and where people kind of have some established ground rules that they can participate in. And so starting from the ground up with that in mind and with that ethos and putting that out there, into the world and locking in on that vs. Google, for instance, is so extremely powerful, they just really don’t care what anything, or what happens within the ecosystem to some degree.”

Kohji: “I imagine as a consumer even, it’s probably a little bit dangerous if that sort of thing is in the wrong hands, because you are only going to see what they want you to see, as opposed to what you actually want to see. You don’t even know what you want to see, but you only know what they're feeding you. And especially in the wake of all this internet privacy business, it’s probably a big fear when a company like Google has become synonymous with search, kind of like Kleenex or Q-tips or something. So, how are you guys going to differ from that? What makes me want to use Presearch, how is it going to be more fair?”

“…there is a fairly significant demographic of people that are starting to become more aware. and they are more concerned, so we are taking a pretty strong approach to protecting people’s privacy.”

Pape: “Sure. So, from the privacy standpoint, the transparency, there is a fairly significant demographic of people that are starting to become more aware. and they are more concerned, so we are taking a pretty strong approach to protecting people’s privacy. Giving them choice and control over their data and how they access the platform. Giving them lots of ability to customize the output and so rather than this passive personalization approach that Google is taking where they observe what you are doing, which leads to them tracking you on all these different devices and across all these channels, ours is more about enabling you to specify your preferences and what you want to see which could include the results that you see as well as potentially the different demographics that you might want to support. For example, I want to support local businesses, I can put that in as a filter and then rather than get a whole bunch of results that are all big-box stores or Amazon, I can get local businesses that are around me. And so there is that approach to things, we also reward people for searching. So, basically [you]get paid to search. You can earn Presearch tokens when you switch to the interface and then we really have been born from this focus on web-workers, people that are power users, people that are on the internet all the time. And people who want to be able to not get this ‘one-size-fits-all’, ‘vanilla search’ output. So they can go in and they can specify different resources that they might utilize. So maybe you are doing domain name lookups all the time or if you are looking into any of the thousands of different resources on a regular basis. You can basically have those instead of being multiple clicks away or you have to create this big bookmark menu or whatever, you can basically add them in as providers into the Presearch ecosystems and then deep search those resources directly from one search field. It even works, for instance with some Google apps, more easily that Google. So if you want to have one click into Google drive or into any of the different Google resources it is actually a faster way to access those as well. Because Google is really targeting everybody, their stuff is fairly generic. This enables you to be very specific. [Presearch] actually saves time and is more convenient.”

Kohji: “So it sounds fairly robust and complex in a good way. Where was this idea born from? Cause I can’t imagine you work up and day and were like ‘I’m going to do all these crazy things.’ So where did you come up with the idea for Presearch?”

“…we saw this creepy dark side of how much power they [Google] have and how there was no dispute resolution process.”

Pape: “So part of it was from the experience we had through our other company, ShopCity.com. In 2011 we woke up one day and Google had demoted all of our sites. There was a penalty that was applied. So we had a site, for instance that was online for almost a dozen years, [with] all these local businesses participating. You wake up one day and the site, even if you Google the name, is on page eight. And then, [we were] trying to get through that incident where we had to participate in a US senate inquiry into Google from an anti-trust perspective with the FTC, with this lawyer. In order to really reach anybody and actually get any results, we saw this creepy dark side of how much power they have and how there was no dispute resolution process.

So there was that and then just being web-workers ourselves we’re online 20-hours a day some days and always accessing these resources within our workflow and just finding it to be inconvenient. And realizing, ‘Hey, if we built this system (so we did this in 2013) we made it faster and easier to search those.’ And if it had value then that product just becoming very sticky and lasting and standing the test of time of 4 or 5 years. Just being very convenient. And there was an opportunity to combine the two and then really, leverage the crypto-space and really, more than anything, trying to create a really thin wedge in user behaviour. So you can actually still search Google through Presearch which makes it very low-switching cost. So if we can start to get some of that search behaviour and search attention through the platform then we can more easily put in different services and enable people to access different things of their choice.”

Kohji: “I image that most people are actually itching to get on the first page of whatever search they are doing. So how do you prevent what happened to you, happening to someone else? Is it democratized in some way?”

Pape: “There is really two phases of the project. We launched a beta on November 9 [2017]…”

Kohji: “My birthday!”

Pape: “Oh, that’s a great day. Fantastic.”

Kohji: “Amazing day.”

Pape: “I knew there was a reason why we picked November 9.”

Kohji: “There you go.”

Pape: “Between and then and the beginning of May we signed up about 13-thousand beta tester. And then between May and now (June 5th) we’re up to over 150-thousand beta users. And so on that platform, it was really about leveraging existing search resources, enabling people to search Google, Duck Duck Go as well as about 85 different providers (Amazon.com) that can be found within the system, while we build out the search framework for version two. Which is a lot more complicated, there is a lot more involved. We’re tyring to figure out the governance model. We’re trying to figure out the spam prevention model, we’re trying to figure out that dispute resolution model and have all of the technology behind all of those things. So we’re looking at a fundamental paradigm shift from being a search engine and, ‘Hey, we deliver a search result where we say, ‘Hey, here is the best thing.’’ We’re really determining relevance, we’re really trying to provide a framework where other people can participate and other people can decide [what is relevant]. So you may have, for instance, we’ve been speaking with the University of Waterloo [Ontario, Canada]. They have an awesome computer science department. They have all these algorithms that they’ve created over the years that they have never really had a way to monetize and to really get large scale access to usage. And [we’d ]enable them to create an algorithm and feed it into the system for usage and testing. Enable subject matter experts to curate collections of information that they have specialized knowledge around and enable all these different participants within the ecosystem to get access to users and to then earn tokens and share in the traffic and monetization of the platform. So it’s really about creating that ecosystem. And so from that spam prevention, there is definitely a significant focus on user feedback. We’re looking at different models like staking tokens and enabling people to establish reputation and trust by staking something of value. It could be our tokens, it could be other digital assets into the ecosystem so that they can establish something that can be trusted and potentially that even could be lost. There is a model there where we could enable people to stake something and then if enough people say, ‘Hey, this is spammy,’ they might actually lose that asset. So looking at all these different models, it is all kind of new, but we think that rather than take the approach that Google has taken, where they basically just say, ‘There are so many bad guys that we have to hide everything from everybody.’ We think that there is a way to really solve it more by being open and establish those ground rules and set a level playing field so that everybody can participate equally.”

Howitt: “It’s kind of interesting, spam has been the focus recently. And that came out of a conversation with one of our potential partners. They were discussing how spam is being indexed by Google to a point where the spam is more popular than the brand. So when people search the brand, the spam comes up first. And now the brand behind that is saying, ‘I don’t want anything to do with Google because it’s actually hurting my brand’. So that is where Colin’s impetus is for that on the spam [issue]. It’s because they came to us and said, ‘Hey Presearch, let’s do some pilots and figure out how can we avoid these things’. And some of those ideas were fantastic.”

Kohji: “I know that you are still kind of working this out, but is the idea that the users of Presearch will determine what is spam and what isn’t?”

Pape: “This is really one of the fundamental shifts as well. There isn’t going to be one index. There will be multiple indexes and people will be able to control which index they access and which filters are utilized. It’s going to be A) a lot harder to ‘game’ the system because there isn’t going to be that one data set. For instance, there may be a organization that has either a filter or a data set that you might have an infinity for or you might want to support. And you might choose to support them. From a spammer standpoint, it’s going to be a lot harder to pollute some of these different indexes because there isn't just one. Then [we are] looking at leveraging something like Reddit, with upvoting and karma. Enabling people to curate and upvote and downvote the information rather than have it all be done centralized. Then have, as part of that process, a tribunal. So let’s say something happens, like us, and you get wrongly identified as spam. Provide people a way of going to the community and they can petition their case and then maybe get a resolution rather than [hitting] a brick wall.”

Howitt: “That’s what happened with Shop City. There was no tribunal, it was into the void.”

Kohji: “Right. Well, I mean, necessity bred invention, right? Ok, let’s talk about the part that is most delicious to me. How do I get paid, to search? Because I like that. I search all the time anyway. Yeah, let’s talk about the token part of this whole thing.”

Pape: “Yeah, it’s the PRE token. P-R-E. And you can earn them, you go to Presearch.org. P-r-e-s-e-a-r-c-h dot org. Create an account. We really don’t require much information at all. Then you can go in and you can customize the providers that are within the ecosystem. By default, it’s currently searching Google and there are a bunch of other engines. And we include any platform with a search results page as being an engine. So it’s not just a generic web search engine. We do have Duck Duck Go, we do have Bing and others. We are in the process of launching the alpha version of the Presearch engine as well. But it could be…Facebook has a social search engine. LinkedIn same thing. Amazon is the number one product search engine in the world. So you can customize which ones you use and then over time we’re going to be adding all these different options. For instance if you’re using Trello or if you’re using any of these CRM systems, you’ll be able to deep search into those. And then we’re shortly releasing the ability to customize, so that you have the ability to add a platform in without waiting for us to do it.”

Kohji: “Amazing. As soon as you guys get AltaVista, I’m in.”

Pape: “Yeah, you just search and you earn tokens. We have a bunch of advertising and sponsorship that we’re launching as well, we’re going to enable people to earn tokens for interacting with brands as well.”

Kohji: “Is this built of Ethereum? Or…”

Pape: “Yeah, it’s an ERC-20 token.”

Kohji: “Ok. So let me ask you this.. We’ll get back to Presearch in a second, but now that we’re into the whole crypto-space, are there any crypto-projects that sort of peaked your interest? When you were going through this whole thing and deciding how to build it out. Where there any that caught your eye? Or just personally, ones that you’re really into?”

Pape: “Yeah, I think along the same lines Steem, Stemmit [are]really interesting models. Seems to be pretty successful and working well. BAT as well (basic attention token), they’ve got the browser and they are basically enabling you to support the projects that you want to support through their token rather than seeing advertising, which is really interesting. Personally, I really like Dash. I think Dash is an amazing cryptocurrency. They just seem to be very well aligned with my personal ideals, from a liberty and freedom standpoint. And they’ve got some amazing tech behind them. I feel really strongly that they should get a lot of recognition. Those are some of my favourites.”

Kohji: “Awesome. Now let’s talk about Canada in general. It seems from my perspective, maybe I am a little biased doing this show. But it seems to me that Canada is on the forefront of the whole Crypto movement. There is all these, I mean, you guys, and Ethereum, and all these other people that we have spoken to, and hopefully will speak to, fingers crossed, that are leading the way in the crypto-space. Why do you think so many Canadians are adopting crypto, or they think it’s the way of the future?”

Pape: “It’s pretty awesome, especially with Canada being considered somewhat of a conservative country from a start-up standpoint.”

Kohji: “Yeah, we are kind of risk adverse, this is new territory for us.”

Pape: “It’s really exciting, definitely Ethereum, I think, has had a huge impact. The creators being here and all the leverage that that has given the space; capital, talent. I think has been tremendous. I have always kind of theorized that the winter months, honestly, really lead to a strong development community. We’re stuck indoors for 7–8 months a year and it breeds a really perfect environment for coding, hardcore.”

Howitt: “I thought it was for gaining weight.”

“Toronto in particular, being such a hotbed for technology and cultures from all over the world and being a bit of a melting pot, I think it’s the perfect place for crypto to really thrive.”

Pape: “That as well! I think, beyond that, what is really exciting for a lot of people I know. Definitely for myself, is really this notion that crypto is this tool to level the playing field and whether it’s Democratic, or something along those lines, how people feel about it. It’s giving power to the little guy. It’s giving an equal voice to those who may be marginalized within the existing system that’s been established for so long. And has a lot of gatekeepers in place to prevent a lot people who, I think, are really well-intentioned and creative and have some huge value to offer the world from becoming capitalized and funded and supported. And I think that really jives well with the Canadian ethos of trying to be really fair and just and wanting to have a level playing field. So I think from an alignment standpoint, it makes a lot of sense. Toronto in particular, being such a hotbed for technology and cultures from all over the world and being a bit of a melting pot, I think it’s the perfect place for crypto to really thrive.”

Kohji: “So I feel like you’re trying to do something that is a little bit, not outside the norm, but is definitely interesting and a little subversive. Have you, has there been any pushback at all from people, [who are] afraid? A lot of people, up at the House of Google, they’re probably listening to us, right now, through my phone. You know?”

“…we’ve attracted people who are ‘f-the man’ attitude, and that’s our earliest demographic of supporters…I think that’s really powerful. And we are just looking to empower other people who share that same viewpoint.”

Pape: “It’s a consideration. And that right there. The fact that you feel that. I feel that. And we’ve had so many people say that, that right there speaks to why it needs to be done. And we definitely felt that in the earliest days when we were looking at how can we potentially get Presearch out to the market. You [would] talk to a venture capitalist and they were like ‘Yeah right. There is no way we’re doing that”. And even with ShopCity, when we had that issue, we were trying to get press, trying to get people to cover it and they were like ,“Yeah, no we’re really not going to risk getting blackballed by Google, but good luck with that”. Until we finally found someone that broke the ice for us. I think there is a little bit of that sense right now We’re probably too small. I know they’re probably aware of Presearch in general, but at this point [we’re] probably not a threat. And truthfully, they stand to still participate within a really high level within that eco-system. If it goes to the next level. I think with the whole crypto space, there is a whole element of subversion that I think is really attractive, truthfully, for those same reasons we just spoke about. A lot of the systems, they’re flawed. And it’s the gatekeepers that are parasitic. They’re extracting value, they aren’t creating value. That are being rewarded. And I think a lot of people are just hungry for alternatives and something that they feel like they can access this system without having to get permission. I can create and be rewarded without fear of being marginalized and with us, I think we’ve leverage it. We confronted it right out in the open in our whitepaper and our marketing and we’ve attracted people who are ‘f-the man’ attitude, and that’s our earliest demographic of supporters. Fortunately, having those supporters gives you some strength, they’re not that lone individual espousing some views that are maybe a little bit controversial or threatening to the powers that be. I think that’s really powerful. And we are just looking to empower other people who share that same viewpoint. We’ve been doing that already through the project. Connecting with people. That is actually why we like Dash. Dash is definitely along those lines. It’s interesting.”

Kohji: “Those people have loud voices. So I'm sure if they’ve got your back a lot of people are going to hear about it. So the one question that comes to mind for me, is if people can create their own indexes and fine tune the search to varying degrees is there any fear that some people might get caught in an echo chamber of only getting the results that match their narratives as opposed to the results that they maybe should get. Are a bunch of people going to start using this and think the earth is flat?”

Pape: “Yeah, it’s the new flat earth index. It’s definitely a potential issue. I think what is different, is that at least if people are selecting it they’re not getting an algorithmic filter bubble where they don't even realize it. We are looking at ways that we maybe surface stuff like, ‘Hey, this is one of those search terms where there are multiple viewpoints’, and give people a list of different viewpoints. And put that out there.”

Kohji: “I guess in that case, these people are building in their own ignorance wilfully as opposed to just by accident…”

“We see one way to engage with the community is to enable people.”

Pape: “I think one of the big concerns in addition to the privacy side of things with Google and others at scale is people’s concerns that they actually have agendas, that they [Google] are forcing, and that there isn't really that same free thought that everyone thinks there is. That there is some type of an agenda that is being propagated, and so with what we’re trying to enable to [be] create[d] is that won’t really be an issue. There won’t be one central index, it won’t be “Presearch has a viewpoint, and this is what I’m getting”. There is going to be control. We see one way to engage with the community is to enable people. So let’s say you spend a bunch of time, and get your Presearch configuration really dialled in and it’s providing you with awesome results and you can put out there some of your beliefs and some of the things you are aligned with. And you might be able to share that. You can then monetize it so that people don’t have to do all the same deep digging that you’ve done. It’s a way to compensate the people who take the time. And again, you are putting out there, actively, “Hey, these are the things I support and believe’, and ‘This is what you’re going to be supporting and seeing when you are using this configuration’.”

Howitt: “A simple example would be, I just did 20 hours of research on travel to the Galapagos island and I’ve documented everything and after could go and say how my trip was, the travel company and everything else. And I could share that with people and they could ‘trip-advisor’ rate it and ‘tip’ me for it. Monetize it in some way. It’s curated lifestyles.”

Kohji: “Nice. I like that.”

Howitt: “And that’s why we say, we can index a certain lifestyle. Vegans for instance, vegan lifestyle, pregnant mom lifestyle. So these could all be separate indexes that you can customize based on where you’re at in your life.”

Kohji: “Nice. Nice, thanks. So what is the, what do you see in the near future for Presearch, and what is the long term, maybe pie in the sky, maybe not? How do you see that going? We’ll start with short term first. What are the next steps for Presearch?”

“…out of everybody who’s signed up to be a Presearch user, more than 40% of them have stuck around and become daily active users.”

Pape: “So right now we are continuing to build out our team. We've just added some awesome talent with some really deep domain expertise in the search space and so we're going to continue to build that second version out. The first version is coming along really nicely. We’ve got a lot of the core systems already built out and they’re working well. We’ve had a much better response, truthfully, than we expected. Obviously, you always hope for the best, but out of everybody who’s signed up to be a Presearch user, more than 40% of them have stuck around and become daily active users. Which is a massive number.”

Kohji: “Nice.”

Pape: “We’ve built out, for instance, a Presearch results page. Which is pretty elementary, truthfully. I think we spent three days coding it. So out of all the people that utilize Presearch, Google is the default. We give that to people, and then the Presearch one, they can choose. About 65 % of people are getting the Google results. And 25% are actually choosing the Presearch results page. Which we’re just blown away by. So we are going to keep really refining that model. In the short term, we’re looking to get people signed up onto the platform. We’ve launched a referral program. That’s how we went from 13 thousand to over 50 thousand [users] in under a month. We're really going to step on the gas there, do a lot more marketing, getting the word out.”

Kohji: “I think it makes sense to just start with Google and just give people what they’re used to and slowly wean them off…”

Howitt: “It’s quite fascinating. I have a 15-year old son, and he’s only ever used Google for all his school work and so on. And so obviously in our house, we’re Presearch. So he’s got his Presearch search bar and he’s got Google, but he’s got all the search engines he never knew about before. So now he looks up his school project and he clicks Google and he gets his answer. Great. And he goes and he clicks Bing. And he goes, ‘Wait a minute, that answer is different.’ Then he clicks Duck Duck Go. Different again, Then he clicks Twitter. And he’s getting all these different answers and he’s like, ‘What’s going on?’ Well, Google is maybe showing you the answer you might want to see based on a biased result or a paid result, but across all these different things, he’s getting better results. So he’s being using [Presearch] exclusively for doing his essays and assignments. So he gets a better sense of what the real information is, and he’s kicking ass, I’m telling you right now. He just got 95 [percent] on his last essay, because he was more complete. Everyone else had Google answer and he had this.”

Kohji: “You’re scaring me. this is like ‘1984’-esque. stuff.”

Howitt: “It is totally ‘1984’.”

Pape: “It is to some degree, for sure.”

Howitt: “But because the tool bar is so convenient to use, you’re doing things faster. You’re being way more productive and you’re getting rewarded. And then everybody is going, ‘Woah’ and we were like ‘Oh’. Like Colin said, we are doing way better than expected.”

Pape: “And long term, creating this environment where people can create value and create and discover information and be rewarded. Its pretty lofty ambition, but really we think than rather than try to be the ones who create the definitive thing, if we can help rough-in a framework and enable different communities to self-generate and coalesce around various components and enable different crypto currencies and different resources to plug in and be connected. Ultimately, where the real value is, is in the brand, and in the usage and in the initial search field that just makes access to all these different things really easy. We could definitely decentralize that search power and create more choice and just ensure that the system itself is in the hands of the community. Its not controlled by any single individual or entity.”

Kohji: “Crush Google under the thumb of the 99%.”

Pape: “Oh yeah.”

Kohji: “If there is an SEO robot listening right now, I’m sorry. I love you Google. I’m just kidding.

So if people want to check out the project, where do they go? I know you mentioned it earlier, but let’s hit them with it one more time.”

Pape: “Just check it out, presearch.org if you want to sign up. There is an info site, presearch.io as well.”

Kohji: “I wish I had a referral code to give you guys [his audience]that I could rake in that coin, but I don’t. Thanks for sitting down and chatting with me. And when things come up, please come back and let us know about it, because I love this. This is great. I’m looking forward to getting some results that I’m not accustomed to because I think I’m as ignorant as the rest.”

Pape: “Thanks for the opportunity, you guys are doing awesome work and we are stoked, one more amazing Canadian crypto resource with you guys. Thank you CryptoCanucks.”

Kohji: “Sign up for Presearch, make sure they index our podcast, the rest is up to you.”

You can listen to the entire podcast by clicking on the link below or download it for playing later:

CryptoCanucks Podcast #3: Presearch with Colin Pape and David Howitt

Presearch is a decentralized search engine that doesn’t collect your personal information. We offer access to over 80 different search providers while rewarding you with PRE tokens simply for searching.

www.presearch.org

Presearch is building a decentralized search engine — visit www.presearch.io or https://twitter.com/presearchnews for more info