Jul 20, 2010
This article continues our 6-part series on semantic technologies bringing you interviews with industry thought leaders. To follow this week are interviews with David Recordon (Facebook), Will Hunsinger (Evri) and Jamie Taylor (Metaweb).
I recently spoke to Chris Messina, a well-known advocate of the open web. He is a board member of the OpenID and Open Web Foundations, and plays an instrumental role in advancing OAuth and safer online computing. In 2008, Chris received the Google Open Source Award recognizing his community work on initiatives like microformats. He is credited with introducing hashtags on Twitter. He currently works at Google as an Open Web Advocateand resides in San Francisco.
I read that just before you took your job at Google, you were thinking of doing a start-up – how does it feel now being in a big corporation like Google?
Well, actually I did have a company, doing consulting for a while and I worked for a lot of start-ups and early-stage companies helping them to craft a community strategy, working with them on marketing and – given my experience in open source – figuring out how to build technologies that then had community belts around them.
So, when it came time to figure out what I want to do next, it sort of occurred to me that if I wanted to do a start-up in the social web space, where I do most of my work, there would be a lot of stuff I would have to build which wouldn’t necessarily add a lot of value to the business but would have needed to be done because the market place hasn’t yet matured. So, given that, it seemed it would make more sense to spend some amount of time at Google, trying to move the industry forward, in terms of getting standardization on a couple of basic features that enable social interaction, working on identity technologies, activities streams, social data formats and things like that in order to create the pre-condition for more interesting type of businesses that perhaps can be created later. In a way it was a sort of like you need to do A before you can do B position.
I think Google is certainly interested in making the web a better, richer place to build these kinds of applications without having people to go and reinvent the wheel all the time. I think it is getting to be a real tension point for consumers where the next start-up that shows up asks them to create another username and password and they are kind of like “I m done with this, I have 40 different accounts already”-
This fits into the vision of the OpenID project…
So the basic premise of OpenID is simply that if you already have an account on one website that should really be good enough and as long as the website you are visiting is willing to accept another website vouching for who you are and can do so in a reliable way, then you don’t need to create another remote account with another password. So that basic idea is what OpenID is dedicated on and the last couple of years it has matured a bit. It has gotten a bit more stable, widely adopted by people from MySpace, Google, Yahoo, AOL and a large number of organizations on the board itself pushing the standard forward.
I think we found ourselves on a pretty interesting juncture where the technology itself is pretty simple in terms of what it does, meanwhile we found that other technologies namely OAuth provide some competition to OpenID because OAuth sort of avoids the whole identity question and just provides the remote site with a token given access to data.
Facebook, by doing that, allows users similar benefits they could have gotten from OpenID, however revealing a lot more data…
Right, I think the basic thing that has been demonstrated over time, by the point you are making, is that OpenID was a technology that grew out of a largely privacy-centric world view, I think, where you only admitted as much information as necessary to complete the transaction. I guess that approach was not economically valuable enough for sites to adopt that technology versus creating another user account on the fly and people had to fill out the whole form. OAuth came along and did not make any assumptions about the amount of data that should or shouldn’t be transmitted; it essentially left that question up to the site providers. What that meant was that sites like Facebook or Twitter could use OAuth as a way of providing more economic incentives to adopt their technologies. And what happened is that we have seen a lot of fragmentation along what I would call connect-style APIs where Twitter has their own version of OAuth, FB has their own, and meanwhile Google and Yahoo did a hybrid of OAuth and OpenID, but appeared to the standards more directly. Their value proposition is less when looked at it in a net-sum way.
What happens if at some point in the future people are again more concerned about privacy and the amount of data they share via OAuth? Could users force Facebook to give out less information?
In some ways, the horse is out of the barn. People are also getting so much more value out of sharing information, the idea of turning back time and going back to a prior area just does not seem reasonable. Also from a site owner’s perspective: why would they start accepting less data?
I also think Facebook does a fairly good job of providing a good amount of control and granularity over how much data you do release. The problem is that people do not seem motivated to actually go and make changes necessary to take back control. So if your question is whether we can go back to this prior time, the question actually has to be: How motivated will people be restrict access to data. The flipside to the question is of course: I want to have access to my friends’ data but I don’t want to share my data. And if you start going back to the zero-sum game with social data you end up in a much worse situation.
What about Google?
The way that I see it in the work I am trying to do at Google is to make the Google account a more valuable account by adopting technologies that enable more data portability. It should then be up to the users, presuming we provide enough mechanisms to give users confidence about what they share; they can share as much or as little as they want to.
Is Google becoming the Personal Data Locker?
There are a couple of things: On the one hand, what I have heard Facebook say in the past is that they don’t believe that there should be only one identity provider or personal data locker – which I admit is an ironical term talking about moving forward as opposed to back- but anyways, they [Facebook] just want to be the best one.
As far as I am concerned Google, Facebook, Yahoo, Twitter, whoever wants to compete at that level; I think the only way that that vision can be realized is that if we do start to standardize on the number of the ways in which these identity providers talk to service providers and make data available.
How does Google work together with start-ups?
My job is presumable in developer relations, and we do a lot of work to find new partners and work with companies that have good or interesting ideas and that are trying to get off the ground and of course we make a number of APIs available. We also provide a bunch of libraries to help developers – large and small – to get off the ground.
So, part of it is that start-ups are able to build pretty good prototypes that demonstrate an idea or concept and then start to include Google data sets whether it’s using reporting OpenID in general or portable contacts or activity streams, what the Buzz platform is build on, and then we can rapidly step up our engagement and coordination with different groups looking for different features.
Now of course, trying to work within the existing protocols is usually the right way to start – but where there are interesting or new opportunities, especially in the social realm I think Google is definitely interested in finding ways of working with people on innovative ideas.
David Siegel showed in his keynote at the Semantic Technologies Conference how Google’s search is inferior to Wolfram Alpha in answering the question about the current temperature in Venice. A fair comparison?
It’s interesting. I guess I would take a slightly different approach to that specific example. In some respect I would almost want to be able to contact someone in Venice who is a friend or a friend-of-a-friend and ask him: “Hey, in the next couple of weeks, what do you think it’s gonna be like?” because he lived there for 10 years and he knows what the patterns are and frankly whether I should wear a sweater or not. The Wolfram Alpha example is a data-based example which is great for a computer but it doesn’t help me pick out what wardrobe or what shoes to wear – I don’t think Google or Wolfram Alpha could provide me with that answer today. But, thinking about it from a social search perspective, we are starting to see Google integrating Twitter streams in search experience, hopefully providing users with more actionable information, providing a number of different opinions, more contextual data. It is certainly something Google is paying a lot of attention to – information that is contextual to the user, not just generic to the world. The examples that they show are starting to show their age, because if you’re doing the search as a non-logged-in user the answers that you are going to get may not be as good as if you were logged in and connected to your social circle.
One question I should ask David Recordon from Facebook?
Let’s see. That’s a good one! Dave and I worked together for a couple of years, he was at Six Apart before when I was independent, we have done a lot of work together since then. It’s been interesting to see the moves Facebook has made. One of the questions I would ask: When is Facebook becoming an Open ID provider?