At our RealTime CrunchUp event today in San Francisco, the first roundtable is entitled “Filtering the Stream: Getting Rid of the Noise.”
The panel is populated by a lot of big players in the space: Facebook, VP of Product Chris Cox, Google, Google Fellow, Amit Singhal, Seesmic, CEO Loic Le Meur, Futurity Ventures, investor/entrepreneur Edo Segal, CrowdEye, CEO Ken Moss, Microsoft, GM of FUSE Labs, Lili Cheng, Facebook, VP of Platform, Bret Taylor, MySpace, Chief Product Officer, Jason Hirschhorn, Thing Labs/Brizzly, CEO Jason Shellen, OneRiot, CEO Kimbal Musk, and Angel Investor Ron Conway. Our own Erick Schonfeld, Mike Arrington, Steve Gillmor are moderating.
Below find my live notes (paraphrased):
First everyone introduces themselves.
MA: So Brett and Chris were at the last CrunchUp did you sign the deal there?
BT: We met there (laughs).
ES: We’re increasingly consuming information through the streams of data. Address the noise problem, how do you do it?
CC: When we started working on Newsfeed we had a metaphor of the newspaper. It was a way to pull in a bunch of source and find what the reader found interesting. But it’s not just the newspaper, because that fully wouldn’t work. We really focused on trying to think about what you tell people first when you talk about a day or week. You focus on the important things first. It’s the right balance between a newspaper and a stream. It’s a big problem for all of us.
ES: Is your approach different from Twitter?
CC: I think it’s a problem we all have to worry about. There’s going to be more and more and more information.
ES: Ron, you can address this – how many companies you work with are on this?
RC: Most of the companies right now are dealing with the macro issues – like real-time search they’re working on the search. But there’s a huge opportunity here to be more specialized. Someone can say something like ‘we do have the best filter.’ Consumers will look for the best filtering mechanism in the next year.
ES: So what is the best filter?
RC: I think it’s UI and some deep intellectural property – and AI and semantics. Huge opportunity.
SG: Lilly you nodded about semantics.
LC: Yeah, we’re all think about time – search for ‘fort hood’ for example. There are opportunities to make it more personal.
SG: Geo will help with that right, how’s Microsoft thinking about that?
LC: First we have to see the data. We have Twitter data, but how are people going to use location data. How do you make that meaningful? Do you use maps? We’re very early.
JH: UI plays a big role in it. MySpace is concentrating on the stream through a media prism – what music and video are you looking at. We’ve had an open graph, so we have different revelency.
MA: He just threw down against Facebook. Did you say we’re better?
JH: That wasn’t my intention.
MA: Will you announce the imeem acquisition in real time.
JH: Not gonna happen. And it brings me great joy to disappoint you.
ES: You turns on status updates to Twitter. How big is that now.
JH: Yep. Twitter is growing greatly, and that relationship is doing well. I don’t know the stats, but we’re pretty big.
MA: Why won’t Facebook do that? Where’s the fear?
CC: I don’t think it’s something we won’t do. It’s something we’re looking at it, we need to do it the right way.
JH: It’s a pulse of pop-culture. It’s an additive thing.
ES: What’s the value of this stream data? Bing and Google cutting are deals, why?
AS: From Google’s perspective, these are the most exciting times for data creation. It’s exploding, it’s exciting. We’ve been thinking about since Google News and now Google Blog Search. The amount of data coming through is just amazing, and there are great things in it. And Google has expertise in filter and ranking it. Time is the biggest component. Your social graph, social circle, and geo information is all key now too. This is a new kind of information. We would love to get as much information as possible. We’re happy about the Twitter partnership. And we’re happy to get more – the more the better.
LC: I want to add to that. The social information is so interesting because it’s two-way information. It’s a dialogue, that’s really cool. I find out news about Microsoft via my Twitter feed before it’s even announced.
JH: There’s a difference too with how interesting real-time is for one user versus the collective.
JS: I agree. It seems like what’s we’re dealing with is a social finding experience. In Brizzly one of the most interesting things is the “mute” funtionality. Maybe you have to follow someone for social reasons, but you can mute him in your stream and still get his DMs, etc. All the companies here are dealing with different pieces of it.
SG: With FriendFeed you did a lot of filtering. Your thoughts?
BT: I work mostly on the Facebook Platform now. So it’s crazy to see how much it’s used by third-party sites. I now start believing in the opposite of aggregation, it’s all about using your social graph to filter now. I think this is more important than the problem of how do we mix information. If you stop thinking about it as one stream, all these products are producing way too much information, Facebook Connect can do filtering for you I think. If you want to read the comics, you read the comics, if you want to read the news, you read the news.
JS: I can’t believe we keep going back to the newspaper (laughs).
KM: Isn’t that Twitter though? You don’t value that.
BT: Sorry, I didn’t mean to say I didn’t value it, obviously I worked on that, but I think it’s hard to create the perfect experience that way. I think something like Lala has a better way of doing its own social experience, rather than just one giant stream. Realtime is a problem that every product has started to solve.
KM: I agree, but it’s the one stream that has built things like this conference. The problem is that when there are streams that aren’t open.
JH: Sometimes the social layer is a weakness too. I found my real-world friends weren’t as into music as I was. I found random people who were more interested in it. That’s why I think Twitter is important.
JS: People think of these networks in different ways too. Facebook is your real friends, Twitter you can find celebs.
JH: And that’s the reason to possibly mix them.
ES: The value of this data is greater when it becomes public?
KM: I think it’s all about the intent of the user. On Facebook you want personalized data. But aggregating data you’re not going to get it from your social circle. My mother is in your social circle. I don’t want to know about her.
LL: If a status update is public it provides more value? I disagree. Twitter isn’t growing outside of the old people like me. He’s 14, he doesn’t get it. He spends his time in Facebook. He doesn’t want it public. He wants it private. I bet there will be a lot of private growing on Twitter – that’s the key to their growth. Most people aren’t like us in the room.
ES: Why not create private groups in Seesmic?
LL: You can, throught Lists. Sadly, Facebook won’t let me at their filters through the API. We doubled the traffic of Seesmic Web in the last two weeks thanks to Lists.
AS: Shouldn’t all this information be available in one place though? Why go to all these different places? That’s the idea behind our new Social Search. You might think there’s not enough social information within your circle, but in the future there will be more information.
LL: But only Twitter is open right now.
JS: And blending isn’t the only way. If we have this discussion in a backroom it’s different than having it on stage.
AS: I don’t believe that. I think blending is important.
JS: So can we check your email inbox and see what you really think of Bing? (laughs)
LC: It’s great to consume info together, but how do you know who you’re talking to.
JH: So the mode you’re in dictates your activity? You’re different?
LC: I think that kind of works for people.
SG: The death of Office for example suggest that there will be a come to Jesus moment when the public and private streams come together. With DMs they are probably checking that before email. Facbeook has to merge those two streams.
JS: I don’t like the cobbler to touch the Salisbury steak. (laughs)
SG: I think you’re suggesting the user isn’t as smart as they are.
JS: No that’s not what I’m saying.
SG: In the age of Twitter, people have a new skill. How to talk publically.
JS: I don’t think they realize they are talking publicly though.
ES: But only Facebook can do the private and public though right? For Google it’s going to be difficult for Facebook to license this private infromation to you.
AS: Yes, these are the challenges we have to address. From a user’s perspective, if data is available to them, it should be available to them (whereever they are). We have to worry about privacy – but it’s shouldn’t be about the data here or there.
KM: Our users are smart, but we have a confusing world. We have to be careful about how we design this user experience. It’s not an easy problem.
LL: I think it’s really about learning that just because we can do something doesn’t mean we should. Like Paul Carr’s post the other day – what should we really be posting?
MA: But you’re like the worst example of that. Everything with you is public.
LL: You’re a very special case in my social graph. I think we shouldn’t post the location of your house on Foursquare just because we can.
Edo: The best chance to solve the problem isn’t the people on the stage, it’s the people in the audience or watching this outside. Geolocation is the next great domain. And someone not here probably won’t do these new things. We probably won’t experience this all through a search box. Twitter giving Google and Microsoft means they have to give it to everyone – or it stiffles information
ES: So was that a mistake?
Edo: I wrote that post being provacative. But I do think that. You need a jam session to get to that place, and in Twitter they aren’t jamming yet.
RC: Who knows what they’re jamming on. I hope that it’s search.
Edo: It’s important to giving this data to everyone.
MA: I think these things are evolving so quickly that it’s hard to see into the future. I would love to see Google get along with Facebook. And Facebook get along with Twitter. There are a lot of jealousness which they publicly deny, but privately confirm.
Edo: But it’s all about money, right?
MA: I don’t think that much about private and petty jealousness.
JH: I agree with Mike. MySpace wants to work with everyone, even with the press pits us against each other.
EM: We’re competitive. That’s what makes us great. This isn’t France. (laughs) Sorry Loic.
SG: The people behind of course they want to promote openness.
JH: Sure. That’s a part of our comeback. But remember we have our own users too. The Twitter thing has worked for us and them.
SG: So is Facebook and Google going to get together.
BT: I’m not in the deal side either. But we’re do believe in openness. There are apps that publish to Twitter and vice versa. The elephant in the room is privacy for the users. The data is open, not in the firehose way, but you can get access to that information. But how you make that avaiable to a search engine is much, much more complex. If there were an easy answer clearly we would do it. It wouldn’t make sense not to – but it’s a difficult problem.
SG: How long does it takes tweets coming into FriendFeed. Like 30 minutes. Is that a technical problem?
BT: We’re moving to a new API and there are some roadblocks.
SG: The blockage is an economic blockage, not a technical one.
ES: Is realtime search more for navigation?
JS: Potentially. But it’s important to note that everyone here not Facebook or Twitter we’re all reliant on them. I don’t envy their position, it’s hard to listen to all of us.
ES: So what do you guys want?
LL: I want to Friend Lists from Facebook. It’s a pain to do both Twitter and Facebook lists.
BT: Seems reasonable. We should do that. Not sure we’re working on that, but we should be.
SG: Ron what are the big issues third-parties are dealing with?
RC: I think consumers don’t realize how to use all these different services right now. Usability, filtering are really key. We invest in sentiment engines, four realtime search engines.
(Mike takes Ron’s list of investments to laughs)
RC: These new techs like geolocation and semantics, people are going to make great things with these.
MA: When you write ‘stealth’ next to a company what does that mean (laughs)
SG: So Lili, Ray Ozzie brought you in to deal with some of these realtime things with Microsoft.
LC: You have to keep people who are putting in the information interested in doing that. How do you work around walls? What’s the right openness that you want?
Q: Should any of these companies look bigger trends? We have all these tools now, should we be using this stuff to help the world?
RC: I think society is radically changing behavior – and that’s why companies we’re talking about exist. It’s so radical that consumers can even keep up with it. People don’t realize some of this stuff is public.
JH: And I think what Ron is getting at could speak to what you ask. I’m not sure all companies are thinking about this, but we’re getting to a place where many of these companies are changing human behavior.
AS: People used to wait a few hours to get search results. But now everyone wants information right away from any of the search engines. Society will change again too. My kids will demand much more of my search engine though. That’s the fun of real-time.
Q: Thanks for all you guys being here. The number of people who can publish to the web has really expanded – but what about the difference between what I know versus what I’m saying? Should it be sharing about everything I know?
ES: Sergey said that – that he wants to think something and it to appear.
AS: Yes, it’s getting easier, but there is still a bottleneck. Not everyone says everything they know. It takes too much effort to share still. I think many companies will be working on that. We’re moving in the right direction, but there is still a bottleneck.
Edo: But you’re asking what we’ve done in the past. Geolocation is not about you writing something, for example.
JS: It’s also not true knowledge as well. When I worked at Blogger back in the day, people thought all this sounded like knowledge management. I think as we’ve made the text box smaller, it’s easier to express this knowledge. I’m not sure a brain probe will help.
CC: It’s difficult to get in someone’s head like we’re saying. But again, it’s surfacing the content that is critical. Like, retweet, comment, those are all great for bringing up content.
SG: What are the terms of usage of Twitter data?
AS: I don’t know. Not sure about length of time.
LC: Not sure either.
Q: Will search become obsolete – will get get results based on what we actually do?
AS: Search already uses a lot of context – like our peronalized search. And the new social search changes it further. This is a good question. Search will have to transform itself to get more contextual information. Search will become much more personalized to you.
ES: What’s the most promising from an economic point of view?
RC: It was very interesting to see Dick Costolo so much promise for Twitter next year. There are a lot of things, analytics, followers, etc.
KM: Search is where there is gobs of money to be made. The problem is how do you predict for the advertisers. When Sarah Palin talked about Glenn Beck, there were no great ads for that.
JH: I think some of the things about geolocation are interesting. If you have an always on device – if you walk down the street and there is a deal.
Edo: It’s the future, we’re just using the wrong word. Ambient streams is the right word.
RC: And discovery is a huge, huge opportunity. The IP has to be perfected. In 5 years this is a multi-billion market.
JS: It’s hard to make an assumption for the best monetization right now.
BT: When social networks took off, every product on the web became social. I think all the verticals of products will start incorporating realtime elements. That will one of the biggest impacts.
Edo: Let’s come up with the right vocabulary. Talk about it on Twitter. We need a good vocabulary. If it’s not just search, what is this?
Video: Hosted by Ustream
Transcript: Provided by PLYmedia.
Come up here, find your nametags, find your name and we’ll sit down.
So this is going to be roughly a 90-minute roundtable.
And we hope to set the stage for the rest of the conference and really sort of touch upon the
Themes that we’re going to be drilling into for the rest of the conference.
Just a few points of sort of order in terms of what you expect to see today.
We’ll have this round table.
Then we’ll have another Q&A with Marc Benioff who had a very interesting announcement yesterday
At dreamforce or two days ago about a new product they’re doing.
Then we’re going to have the product demos.
Before lunch and after lunch.
We sort of broke them up.
And after lunch then we have a geo stream panel.
We have a media streams panel.
We have a conversation about how e-mail is becoming less and less relevant, with Paul Buchheit,
Started Gmail, and Rob Goldman one of our TechCrunch 50 companies and in the end we’ll wrap
It up with sort of where the rubber meets the road panel.
So is everyone sitting down here?
I think what I’ll do is just have everybody introduce themselves, I’d like everyone to give
Me the Twitter version of who you are.
Just quickly go around so that everyone knows who’s on the panel.
And then we’ll just jump right into it.
>> Jason Shellen, co-founder of Thing Labs.
And we make something called Brizzly.
And it’s open and free for everyone today.
>> Kimball Musk, CEO of OneRiot.
Real time search engine.
>> Bret Taylor, director of product and Web set Facebook for the Facebook platform former
Ly CEO of Friendfeed.
>> I’m Loic Le Meur and I’m with Seesmic, one of the clients on Twitter.
>> Chris Cox, head of product at Facebook.
>> Lili Cheng, I run a lab at Microsoft that looks at social experiences.
We’ve been working on a lot of the Twitter search stuff.
And also Outlook stuff.
>> Ron Conway, Angel Investor, love the real time data segment and have invested in about
20 companies since last June.
>> Amit Singhal, I’m a Google fellow I work with the Google search group, have done that for
>> Jason Hirschhorn, I lead product at MySpace.
>> You’re the chief.
>> It’s funny, Chris and Bret, you guys were at our last CrunchUp and that’s where you inked
The deal for the friend deal, is that right.
>> That’s the first place we Ed.
>> Edo Segal, Futurity Ventures, I’ve been thinking about this problem for quite some time.
>> One of the big folks of this roundtable we’re seeing all sorts of new streams of data enter
Our lives from Twitter to Facebook to MySpace, Google Wave, you know you name it.
The way we’re consuming information increasingly is through the streams of data, which has
Its benefits because you’re very in the moment.
But also you sort of become inundated with information.
So I’d like just the ability for people to keep up. So what needs to be done to make this
why don’t we start with Chris.
>> Chris: So, sure, we have this metaphor of the newspaper.
That understood from the reader’s perspective what was important and what was interesting.
And to imagine if the newspaper were comics and all the headlines were interconnected, that
There’s a balance to strike and the flow of everything that’s happening as it’s happening right
So we really focused on trying to think about when you tell somebody about a day or a week,
You start with certain events that are more important or more rare or more germane.
And you focus on people.
On people that are more important to you.
So if something not that
you might care.
If it happens to a stranger you might not.
Whereas if there’s a very rare or important or monolithic event, like an anniversary or a we
Dding or an engagement, that’s something that you care about even if it’s a stranger.
And so we really tried to look at aggregating context around actions, trying to understand
Who is important to the viewer.
And trying to understand which events are more or less rare.
As the basis for building a product that’s the right balance between a newspaper and a stream.
And I think those are the questions that we’re all going to be needing to come back to as more
And more information is flowing, where the attention we have in a day is not going to increase
>> And do you see sort of your approach being vastly different from Twitter, from other people’s
Approach to where Twitter really is at this point is just this undifferentiated stream, isn’t
>> Chris: I think it’s a problem we’re all going to need to approach.
Just because like I said, as five, 10 years go by, there’s just going to be more and more information
Flowing in real time.
And you’re not going to have 24 hours a day to sit and watch it.
I think we’re all going to be approaching this problem.
>> Ron, you can address this.
How many companies you’re investing in are trying to address this problem of filtering the
>> Ron: I think right now most of the companies are dealing with the macro issues.
Like if it’s real time, the search, but there’s a huge opportunity yourself by making the search
And you’ll see the companies start to get more specialized and differentiate themselves by
these because consumers will absolutely look for the best filtering mechanism in
The next year now that they’re getting the fire host.
>> What’s the best filtering mechanism?
Is it a UI or real time search.
>> I think it’s probably a UI and some very deep intellectual property and artificial intelligence
And semantics technology that will end up producing a great filtered product.
So it’s a huge opportunity for all the start-ups out there.
>> Lili, you were nodding about the semantics, I think?
>> I think we’re still not very
we’re just beginning.
So I think the notion of searching what’s the proper time interval if I search for fort hood
Today it’s different than if I searched for it a week ago, knowing who I am and my relationship
To the thing I’m looking at.
We don’t really have it in search today.
So I just think there’s lots of opportunities to make it more personal and also just think
About time more broadly.
>> And the geo stuff is basically that’s going to really turn on the floodgate of data.
>> How is Microsoft going to process that?
>> You know, I think we’re going to
first we need to see the data, right?
>> Well you’re seeing the data.
>> You were telling me last night that you’re sitting there happy as clams.
>> I don’t mean
I mean we have the data for Twitter, but how are people really going to
Start using location data every day in their lives.
We don’t really know that, how people are going to share it.
How people want to consume it.
And then as you overlay this information on maps, how do you actually make that meaningful.
Do people want to see maps with information overlaid on it?
I just think we’re early.
And it’s going to be interesting to see how people consume it and how you make the information
Valuable to people.
>> I think UI plays a big role in it, specifically when I look at MySpace concentrating on
The stream, we’re going to look at it through a media prism and what music is going on in your
Network, and what video is going on.
And given the open nature of what we’re dealing with, where unlike some other social networks,
We sort of had an open graph where you were collecting as many friends as possible.
The more relevant stuff becomes what media are people interested in.
How do you auto play list.
How do you auto watch TV.
And some of the new services that you see like redux or Brizzly what they’re doing with media
I think are very interesting to us.
>> Did you see that?
>> He just threw down against Facebook right there.
He said you said
I don’t think you quite said this exactly.
But you said we are open.
Facebook is closed.
Is that what you said Chris, are you going to let him say that.
>> That’s what I said.
It wasn’t my intention but with Mike Arrington here, throw the gasoline on, baby.
>> The main question I have is are you going to announce the Imime acquisition here on stage.
>> In real time?
>> Just do it.
Just do it.
Yeah, we bought ‘em.
>> Not going to happen.
And the fact that I disappointed you brings me great, great joy.
>> Some of those services, I think, are doing unique things.
And I think whether it be the feed or the stream, this idea of settings, when you grow your
Graph, becomes overwhelming.
And I think sort of in line decision making of I want more of this or this will play
>> You turned on status updates in MySpace to Twitter.
So I can opt in to have my MySpace status appear in Twitter.
And you’re a big portion of Twitter messages now, right?
>> I don’t know the exact number.
But obviously Twitter is growing greatly.
But we’re doing well, and I think our link short ners are doing well.
I’d like to see more relationships like that.
>> I don’t know the actual ratio.
But we’re pretty big.
>> Why won’t Facebook do that?
>> Do what?
>> Anyone from Facebook here that can answer that question?
>> We’ve got several people from Facebook who can answer that question.
>> What’s the question?
>> It was an accusation.
>> Why don’t you do proper two-way synching with Twitter.
What’s the fear?
Why are you so scared of Twitter that you won’t do that.
I guess that’s how I’ll put it.
>> I don’t think it’s something we won’t do it’s something we’re exploring and looking what’s
The right way to explore it and right way to manage it with users.
>> They did it.
They did it, the company you bought did it.
>> The way we look, I think long term about Twitter, the way we look around Twitter specifically
When we think about media, pulse of pop culture that’s a point of entry where a pulse is being
When we want to display stuff to our users, it’s an additive thing.
I don’t look at it as being competitive.
>> What’s the value of all this streamed data?
You know, are you
Google, you just both Google and Bing cut these deals to get the stream
Bing did it for both Facebook and Twitter.
Twitter did it.
So explain to me what’s the value of this
>> So from Google’s perspective, these are some of the most exciting times in data creation.
In the last couple of years, the amount of data being created has just exploded.
And that’s good news for the users, right?
We have been building systems like Google news.
We started with six or seven years back.
And when my friend Krishna and I were building Google News, we were looking at how much data
Can we get crawled from the news sites in real time and how do we get it to our users within
>> I’m sorry?
>> How is that even happening.
>> Somebody’s taking over our conference.
>> And then when you were building Google blog search, we started dealing with RSS feeds and
How soon can we take RSS feed in and bring it back to our users filtered and ranked using ranking
Technology, which we know a little bit about how to take billions of items and give you the
And when we put all that together, we got Google news, Google blog search, but we
What the new phenomena is just far outpacing everything else that we have done in the past.
And that’s what’s so exciting for me as an engineer who has worked in search for 20 years.
The amount of data that’s coming through is just amazing.
It has great things in it, and we have expertise of course indexing it and ranking it and filtering
It for our users.
There will be many components to it.
Clearly time is the biggest component.
So when we are building log search, we soon realized that Mike’s posts are great however his
Recent posts are even better.
Clearly the clarity of the poster, which we have built over years, expertise with things like
Page rank, comes in very handy.
The time factor becomes incredibly critical, but now we are seeing two added very cool features
Or signals to this is a stream of information, your social graph, your social circle, and the
And when you put all these together, right, Google, Google by mixing signals and getting you
The relevant information, when you put these four signals together, we are seeing a new kind
Of information emerging for our users, which is far more valuable than many other things out
>> You want to get as much of the data as possible?
>> We would
we would love to get as much information that can be brought over to our users.
>> Would you like to have
>> We’re happy about our Twitter partnership.
We’re getting a lot of good data from there.
And we are very happy to get more information.
I’d love to encourage everyone to work with us.
>> Would you like to get the MySpace updates and Facebooks.
>> The more data that comes into our system, the better we can make it for our users.
>> Make a deal right now?
>> I am not on the deal side of the company.
That’s why they
>> That’s what I should have said to him about the
>> I want to add to that.
I think one of the really interesting thing about having the social information is it’s a two-way
So it’s really easy to like look at information that you’re getting from Twitter if you see
It on the search portal.
Reach in and like you said inspect the person.
Then have a dialogue with them.
And I think that’s really cool, because even if you see a blog post it’s hard to know just
The barrier to actually reach in, touch someone, follow them.
It’s just very different.
So I think how communication and search and the two-way dialogue between what you’re consuming
And the ability to kind of participate more will be really interesting.
And then I guess it’s like you were saying, I think probably for everybody, when I look at
My Twitter feed I often find out news about Microsoft via my Twitter feed from people who don’t
Before it actually gets announced.
Before I see it in my e-mail.
I think everybody has that kind of experience, that the quality of information coming out of
Twitter just by looking at your own friends is better than what you see through your in-box.
Sometimes, and through the Web.
If we can bring that to more people, without them actually having to follow a bunch of people
And know who’s who, it’s really powerful.
So the information is there.
>> I also think it’s important that there’s a difference between a user base and how interesting
Real time is to them, versus the net effect of real time.
So there are services out there that show you, you know, sort of a streaming amount of information.
It’s tough to take in and users aren’t coming to the site every 10 minutes to be able to see
But there is a segment that that is important to, versus the news or information that gets
Up there the fastest gets retweetd and redistributed, that becomes part of what gets read first,
What gets used and listened or watched.
And it’s important to distinguish between the net effect and what user base is actually interested
In seeing on their own.
>> I agree with Jason.
It seems like what we’re still dealing with is almost like a still in a social problem stage.
So the stream data is valuable to you if you’re a company and you see your TechCrunch article
That you’ve written has 200 retweets.
But really one of the most popular things that we’ve done within Brizzly is adding that mute
Which really doesn’t make sense, right?
If you’re picking people in your own stream, then you go and follow that person.
But there’s actually a social problem.
The social problem is, well, I need to follow Jason for business reasons but I don’t really
Like following all his updates about his kid’s baseball team or something like that.
Maybe I can mute him but I can still have him pop above the stack and direct message me if
He needs to or get his at replies or something like that.
So there’s that personal level of relevance that is not an aggregate relevance.
So I think when you ask about value, Erick, there’s value to one person.
And then there’s that aggregate value.
So I think there’s room, obviously, all of the companies here are dealing with different pieces
But I think the one that we’re focused on right now is how do we solve some of those smaller
>> Is it an either/or I don’t mute anything, or why not mute him but when other people, when
There’s more signal to his tweets, when a lot of people are retweeting his stuff or there’s
An important message, that actually pops up.
That would be
>> On Tivo if I say I don’t want to see this show I really don’t want to see the show.
>> With Friendfeed, you really delved into this issue a combination of aggregation and filtering.
So what’s your thought on what
>> I actually since I work primarily on the Facebook platform now.
So like I think I’ve seen just the incredible usage of Facebook sort of as an account system
And social graph on third-party sites.
And one thing I’ve started to believe more in is sort of I guess the opposite of aggregation.
It’s sort of bringing in social context and filtering to all the applications that you use.
Because I think it maps more to the way I like think about the world.
Like when I’m on using a product like Lala, having that like large stream of information of
What people are listening to filtered by my social graph, is actually incredibly valuable in
And when I’m on a news site, seeing what my friends are reading right now and having like communication
Tools to talk about that, like in real time, is incredibly valuable, almost more valuable than
The problem of how do we mix this single large stream together and filter it.
Like that’s an important problem.
And it’s an interesting problem particularly for sites like Google, where they sort of
Like a very large information problem.
But if you remove
if you stop thinking about it as one stream and have all these products
Producing more information than has ever been produced in the history of man kind and how do
You actually provide value to your users as a creator of that product, I think that I’m more
And more in the belief of like products like Facebook Connect offering the ability for these
Products to provide like social context and filtering, within specific context of, say, watching
Television, listening to music.
So you kind of go to a product to perform a particular action.
Sort of goes back to what Chris was saying.
Like when you’re reading a newspaper.
You don’t have like a mix of comics and news and like editorial.
They’re kind of split up because you kind of, if you wanted to read the comics you read the
Comics, if you want to read the news you read the news.
More and more as the space matures I think we’ll see more of that.
Like more people focusing on particular types of use cases applying filtering
>> But potentially, Bret, if there was a newspaper, and I can’t believe we keep referring
To newspaper as a paradigm that we want to actually replicate in the future.
If there was that perfect thing for you, maybe it would have, like, hey, here’s comics, a little
Bit of this and that.
>> Doesn’t Twitter do that.
That’s what’s great about Twitter.
You actually get it in
you get it in a stream that is the comics.
It’s the cartoons.
>> I didn’t mean to say it didn’t have value.
Obviously I worked on a product that did exactly this for quite a while.
I think it’s actually
I think part of what I’m saying it’s
we’re very focused on if we
Had every piece of information about every single person you know how do we filter it, it’s
Kind of like we created the problem that we’re solving, too.
And it’s entirely possible that, you know, that it’s very hard to create the perfect music
Experience, you know, within a product like that.
But I have a feeling in a product like Lala has a much better chance of creating a perfect
Sort of social music experience, perfect real time music experience.
And so I’m not saying these products aren’t important or Friendfeed isn’t important by any
But I do think that filtering applies to like, it doesn’t just mean there’s one stream and
We’re all filtering the same stream.
In lots of different contexts and lots of different products, real time is the problem that
Like every product has begun to solve.
And I think it’s very important that like we consider like the platforms that will help these
Products solve those problems.
>> I agree with that.
But so the filtering
so the filtering, the one stream that’s common or the open streams,
I think it’s what’s grown this entire conference.
The fact we’re all here and our companies, is that there’s some amazing bets made that made
This data available and new things like lists and locations.
And others that authority that are going to just let us filter the stream in amazingly different
Ways and tailor it to our products and user scenarios.
But I think sort of the danger maybe, is if there are streams that are pockets of information
That aren’t going to be open, then I’m not sure what that’s going to do to the usefulness of
>> I’d also say that sometimes the social layer is a weakness.
If I take myself.
I remember when I first got onto the Internet, one of the things I was looking for was my real
World friends were not as deeply into music in a way that I was.
And this is way back where you’d post on Use Net and ask questions about stuff.
I found other people I did not know or had access to other interests.
And if you just look at the stream through a social layer, how do you also get stuff that your
Friends aren’t interested in and I think that’s a very sort of important thing.
That’s why I think exchanges like Twitter and other places in terms of general trends are still
>> It’s also the other piece that you mentioned there that is important is that people think
Of these networks in different ways.
Facebook is also thought of as, oh, that’s where my high school friends are.
My friends, real friends maybe are.
Twitter is often thought of is hey I can subscribe to ashton Kutcher or my friend from industry.
And I think MySpace is accused of this as well the different kinds of networks.
So I agree that it is a weakness.
But do they have a certain predilection, and whatever comes next.
Is that going to have some certain bent that’s around business or something else.
>> And thus the reason to possibly mix them.
>> What are you going to
>> I don’t know about possibly mixing them.
>> You’re saying that the value of this data is, becomes greater once the data becomes public.
Which they can make it more valuable definitely for search engines and definitely for outside
I wonder whether it becomes more valuable to Facebook and Twitter as well.
You guys have it so you don’t care but if it adds value to more outside developers isn’t it
More valuable to you?
>> I think it depends on the intent of the user.
If you’re on Facebook intent is to keep up with your friends, read the sports, comics, on at
Twitter so forth that’s where you want personalized data.
But the value of aggregating data is when you actually want to find out what’s going on.
In your social circle.
My mother’s in my social circle.
I don’t want to know what’s going on with my mother but I want to know what’s going on with
Everyone else in this room.
You’re not in my social circle.
>> Your mom just called.
>> You’re saying something very interesting that we take a statement that if a status is public
Has more valuable.
That’s what you said.
Actually very much disagree with that.
Because I watch that Twitter is not growing outside of the what my son calls the old people
He’s 14 years old and he doesn’t get Twitter at all.
And I love Twitter myself.
But he spends his time on Facebook.
And I try to talk to him and understand why is that.
And he sees no value in taking anything public.
What he wants is really the stream of his friends in private and I think what’s really going
To be interesting is how private evolves compared to public.
The explanations why Twitter is not growing so fast anymore as Dick said this morning, because
It doesn’t have a private growth that Facebook has.
And many people actually are not like most of us in the room.
We link to public and search and expose yourself.
>> Why can’t you create private groups in Seesmic.
>> You can, actually.
You have use lease, you don’t have to take them public.
But my point is, unfortunately Facebook doesn’t let me have Facebook use least.
You don’t call them use least but friend list and the API level.
So you can’t filter Facebook stream on Seesmic.
Either on the Web or on the top yet unless there’s an announcement that you would be happy
To integrate it.
>> But instead of worrying about public and private, shouldn’t all the relevant information
Be at your fingertips whenever you need it, public, private, semi private.
Isn’t that the right solution for users?
All users out there?
That you don’t have to worry about, oh, my friends don’t know enough about music.
The world does.
Why should you think about where do I go to find out those people?
Shouldn’t they all be available in one place?
>> But in this case MySpace.
>> The public and the private, semi private results.
So a small example of that is social search experiment that we’ve launched recently where the
View of your social circle is of available to you on the topic you are interested in, along
Side the best public information out there.
Now, right now that stream is built up of public information because of access control and
So on and so forth.
But in an ideal world, why should I have to think whether my friends know or don’t know about
Schnauzers, the kind of dog I have or the world knows about it.
I should be able to get the best information and information relevant to me.
Personally relevant to me.
>> So if your friends have mentioned something about that.
So the likelihood of that is very low, actually.
>> I’ve been playing with that search.
>> The results.
>> You just described Facebook search.
>> I was upgrading my iPhone, and I didn’t even realize after Google’s social search launch
And I typed in 3 GS my friend Matt had three GS you may not think enough information is not
Available but access controls related to the things do make it harder, don’t get me wrong.
But as the systems become more and more open, the future is going to be towards everyone not
Thinking about where the information goes.
They need relevant information at their fingertips and they shouldn’t be thinking about it.
>> The problem social graph is not open.
It’s only on Twitter.
>> As of right now you’re right that
>> And blending is not the only way, though.
I think, I mean, this would be a very different discussion if we just went back in the waiting
Room area and had this discussion about real time streams.
We’re speaking into a different area.
It’s not speaking, knowing that it’s going to be published publicly, and I think when you blend
Those things, you sort of degrade the experience, potentially.
So right now in Brizzly
>> I don’t believe that.
>> When you blend relevant information for me, if it’s from my friends, right, Matt’s iPhone
3 GS review might not be the best piece of review out there.
It should not be shown to general users out there.
>> So can we check your Gmail in-box and see what you really think of Bing.
I think what you say publicly and what you say privately are two different things.
You can’t argue that every place is the same place.
>> We’re going to let you have the Schnauzer space.
It’s all yours.
>> What’s interesting you take something like Facebook they haven’t blended your in-box with
Facebook with your feed.
It’s really important.
It’s not just a consumption experience it’s an information experience.
I need to know who my audience is.
And some issues it’s a UII make this error in Twitter all the time I’m SMSing I reply and I
Forget to D somebody and it’s kind of an issue.
So as more and more sensitive information gets blended together, it’s great to consume it all
Together but then how do you let people know who they’re actually talking to.
>> Meaning the mode you’re in dictates your activity?
>> Meaning the mode that you’re in would dictate your activity.
So you may do things differently in e-mail and IM versus
>> I think it’s an interesting decision that in Facebook they’re separate.
I think actually people
it’s kind of a weird thing, right, that your e-mail is in one page
And you have comments in the wall on another page.
But there’s a reason why that probably works for people, because you want to know who your
You want to know the communication.
>> The death of Office, for example, suggests that there’s going to be a come to Jesus moment
When these two types of streams, public and private, from the user perspective, come together.
Because ever since Twitter direct messages started to become visible, there was a certain flight
To direct message.
It’s the easiest way.
If you have a connection to be able to talk to someone.
Because they are probably checking their Twitter feed before they’re checking their e-mail.
So this is
this is a collision that is going to
what Facebook is doing right now, they’re
Going to have to figure out, in my opinion, how to merge those two streams.
So it may be an interesting tactic that they’re using right now, but I don’t think it’s going
To sustain itself.
I’m sure you don’t agree with that, Jason.
>> Which piece of that don’t I agree with?
>> That these things are going to be commingled as opposed to being separate with
>> I don’t like the, in the Swanson’s TV dinner for the Cobbler to touch the Salisbury stake
I like them separate.
I guess they’re different social modes you’re in as a human being, let alone those facets on
And I think those need to mirror each other.
There’s not one friend of mine who would say the same thing on Facebook who would say the same
Thing on Twitter without knowing up front, okay, I’m synching this from Twitter to Facebook.
>> I think you’re suggesting that the user is not as smart as they really are.
>> That’s not at all what I’m suggesting, Steve.
>> I would think that you wouldn’t agree with me, but that doesn’t mean that that’s
In the age of Twitter have learned a new skill, which is how to talk publicly.
>> I don’t think most of them know that they’re talking publicly, actually.
And I don’t know that it really matters in aggregate.
So for most people, I think if you look at the amount of people that they’re subscribed to
And the number of people who subscribe to them, is very low.
And that’s fine.
And that’s probably good for their social group.
But I don’t know that they know that it’s really public and out there all the time.
I’m not saying they’re not smart.
The comment economy and you had all these news articles about look at these kids, they’re posting
Photos of wild parties.
They didn’t know.
>> But this whole issue of mingling
your point about mingling what your private social
Group is saying versus the public statements, that in fact is kind of the direction that Facebook
Is going, and only they can do it really because they have the private and the public right
When you do search, you get your friends results.
But once you have enough, everyone opt-ins, you can easily surface stuff from everyone on Facebook,
The issue for Google is that it’s going to be difficult for Facebook to license that private
Information to you.
I mean, only searchable information really if you guys do a deal is going to be the everyone
you’re at a
to the extent that the Facebook data is interesting.
>> These are indeed going to be the challenges that we, as we the thought leaders in this
And, yes, we have to respect privacy, because it’s incredibly important that no private information
Be linked to anyone else.
But for me, my private information should be available, and I shouldn’t have to worry about
Whether this is here or sitting there.
>> So one of the biggest problems we face here is our users aren’t
they are smart.
But we’re starting to have a confusing world here, right?
Just on Facebook, how many people who aren’t sort of expert Facebook users are typing in the
Wrong box and posting to something embarrassing on a wall that they meant to be private.
and if we start, if we start adding the now some data is public-public versus just public
To your friends, versus private, to a single person, we’re going to
we have to be really
Careful about how we design this user experience.
And today it’s simple, right?
Twitter, it’s open to everyone.
And Facebook, it’s my friends who I’ve shaken hands with and we’re friends.
And I think that we’re going to have to get to this world where the information is available
So everyone can take advantage of it and users can get good tools.
But it’s not an easy problem.
>> I think it’s
I think it’s really about learning that it’s not because we have those tools available that
We should do everything we want with them so it is if you look at the Paul Carr’s post about
The shootings in Texas 10 days ago, where he actually posted this video of someone dying, and
That’s on the video.
It’s like I have a camera on your face, and like I think Paul wrote: Watch me watch you.
And it’s like
>> I’m sorry?
My point is we shouldn’t post those things.
The guy who was the military started Twittering about the events shouldn’t have done it this
Way, not because he could not do it, we can send it to Facebook, to MySpace and Twitter, but
It’s more about learning the basic effects like if I go in your house
>> The worst example of the Paul Carr article.
There’s no barrier between your private and public lives that I can see.
And anything you hear or do privately, it seems like you’re perfectly willing to push that
Because that is sort of what you do.
I’m not judging you on that, at least here.
>> If I go to your house, I’m not going to
>> You do that all the time.
>> Luic I’m having a bad day.
Heather and I got in a fight.
She’s mad at me the page is down.
I see a Twitter.
Erick’s bummed because Heather yelled at him.
And I don’t think in your brain you see any difference.
>> You’re a very special case in my social
>> But, yeah, I think we have to think about not going to your house and posting the location
Of your house on Foursquare.
I think it’s not something we should do.
It’s not that I can’t do it.
I shouldn’t do it.
>> You wanted to make a point.
>> I think going back to the original question and the context of the panel, it’s very unique
To have all of these people together and I want to just shine a light on a few areas that we
I think the greatest chance to solving the problem is actually not the people on the stage.
It’s the people in the audience and watching us at home in terms of the ability to actually
Break through on some of these issues.
So because innovation tends to happen elsewhere and it happens outside big companies.
It’s very important that the bigger companies are very cognizant to that to the extent there
Remains an openness that kind of Twitter really started.
And Facebook is moving in the direction.
And geo location is the next great domain.
Because the reality is that we have a lot of talented people that are approaching different
Slices of the experience, but the real break through, where the real future is of this, we’re
Actually not even
we don’t even know what questions to ask at this point in terms of how
It’s going to be consumed, because we’re kind of assuming it’s going to be consumed in the
Same way, to your point, about where we are now in a search box.
But it’s probably not going to be that way at all.
It’s going to be much more of an ambient experience.
So my first point is in the day when Twitter gives their feed to Microsoft and Google, it’s
Probably the day or soon after where everybody has access to that feed or otherwise it really
Does become a point in history where it’s actually stifle ling innovation.
So that’s one point.
And I think if Twitter’s point on that is we can’t support it because we can barely keep our
Servers up, then guess what you just gave your data to someone that actually knows how to do
>> You think it was a mistake to do the deal with Google?
>> I wrote that post about it being a mistake with full knowledge that it’s happening.
And I was being provocative, and I think the point for me right now is that it goes back to
What I was just saying about innovation.
It’s kind of role playing.
If you and your spouse
there are certain things she does, certain things I do.
And we kind of fall into place.
If Twitter has basically outsourced its search problem to the best minds in the world in Google
Or Microsoft to do that, they’re not as focused on that list of 20 things that they’re doing.
And that really is a shame, because I think the search and discovery problem is one where you
Need that kind of passion and that
it’s kind of like jazz.
And the right people to have that jam session is the people in the Twitter building.
And right now they’re not jamming on that.
They’re jamming about on boarding.
And that takes them
>> Who knows what they’re jamming on.
>> Some of them are.
>> You do.
>> Do you know what they’re jamming on.
>> I absolutely don’t.
But I hope they are jamming on search.
>> I would argue they’re not.
Or not as much as they would have.
>> I argue that I hope they are.
>> I argue the same thing.
>> That was my whole point.
>> In fairness Dick said on boarding search and discovery.
So it looks like they’re jamming.
>> That was my concern, just about that.
but I think the point that happens really important to give that level of access
To the people in the audience and all those wise developers.
Because it’s a question of talent but it’s also a statistical question of you don’t know what
Sticks and you’ve got to have as many experiments as possible.
That can’t happen without transparency and data throughout all silos it’s not just at which
Twitter and Facebook but it’s all locations we’re seeing that trend it’s critical for all companies
>> The thing that bothered.
I tried to pick a fight with Erick, but he didn’t bite.
I think these things are evolving so quickly it’s hard to see into the future.
I think companies have visions of where it’s going.
It’s so fascinating to me to see how things evolve.
What I’d really like to see is Google get along with Facebook so that we can have access to
That data as well, as Twitter data.
And not just have look to bing for that.
And I’d like to see Facebook get along with Twitter and have full data integration.
>> Throw in world peace, Michael.
>> I’d like to have world peace in our world.
And there’s a lot of jealousy and sort of anger and dislike of these companies between each
Other, which they public deny, but privately confirm.
I would just really like to see some of us move beyond that so that we can have maybe faster.
>> Your problem is a business problem.
It’s a question of where the money is coming from.
It’s the whole
I mean this whole discussion and the filter and the importance of at which
Twitter participating or not, all has to do with the notion of intent and where people are
Sort of who is growing faster.
Facebook is mad at Twitter because they grew so fast they got so much press attention.
And Google this.
Personally that’s what I think.
Because all these companies are doing pretty well financially.
>> I think there’s that presumption but I’ll throw out my opinion.
As a user we want to be open and work with people.
I think the press or anyone else that pits us against each other might have happened here.
I see a world where we work with Facebook, we work with Twitter and any other great technology
Where audiences are gathering, you know, it can lend value.
>> I thought we lived in America.
>> Sounds like we all should be in France.
But we’re a competitive culture.
That’s what makes us great.
The fact that Facebook and Twitter, MySpace compete
>> He was trying to save this for the Web.
>> We actually will be in France in a couple of weeks.
>> The reality, the nature of the Internet is open.
It’s only a matter of time before these systems.
>> But you say that.
>> For me, listen I’ll say it right now.
It’s certainly a part of MySpace’s come back, but there’s also still 100 million people that
Are unique to us, and we have good experiences for them.
But I also think that Twitter showed that all of a sudden it made our stream more vibrant.
And it was a good thing.
It worked for Twitter and it worked for us.
It’s not always I think in that case a zero sum game.
>> Is Facebook and Google are you guys going to get together?
>> Mike wants an announcement.
>> I just want to say like I’m not on the deal side as well.
I can pull that card as well.
I want to say we’re committed to openness.
I think the challenge that Ken alluded to earlier is that there are like different privacy
Controls on all these different networks.
So we did have the ability to sort of like publish your Facebook status updates directly to
There’s lots of applications that do the other directions.
Like I think most people who have that direct need can accomplish it pretty easily right now.
I know my account is set up that way.
For example, and I think most people’s here are.
I think there’s a genuine challenge that doesn’t apply to most people at this table because
We tend to use these services sort of as a broadcast mechanism that do apply to a lot of users
Like whether or not they have the need to publish things from one system and automatically
Have it go to another system.
The implications of that decision aren’t necessarily obvious to a lot of users.
So I think a lot of the elephant in the room really here is privacy for network particular
Ly like Facebook where there’s a lot of private information being shared.
There’s a lot of like implicit and explicit contexts in which that information is shared.
So the data is very open.
Not like in the sort of fire hose way but a user if you use something like Facebook Connect
You can get access to that user’s stream and like Seesmic and Brizzly does, that information,
Private and public, is available.
The question of like how do you like make that available on the search engine is a much more
And so I think that we’re
I honestly don’t know like the answer.
There’s no reason like openness benefits Facebook at this point.
There’s so much innovation on our platform.
That like it would be really, really like strategically bad for us to do anything that would
stifle that vairgs.
But I think it’s a genuinely difficult problem.
>> Before we get back to the privacy discussion, how long is the round trip right now incoming
From tweets to Friendfeed?
>> That’s a little longer right now.
>> Like 30 minutes right now, right?
>> Is that a technical problem?
I don’t think so.
>> We are supposed to
we’re transitioning to a new real time API that Twitter is providing,
And there’s some roadblocks.
But I don’t know all the details.
>> The roadblocks, not to beat up on Twitter or people who are investing in Twitter, but the
Fact is that what the blockage to openness is an economic blockage.
It’s not a technical one.
So when we talk about search, that’s after this war is fought and decided.
>> Probably most convenient figure leaf on that.
And it’s correct.
I think the big war at the end of the day where do these users go to to start their discovery
If it’s Google, then that creates the mapping to the intent which is where most of the money
Is, which is not necessarily where Facebook’s head is.
Where it is in the future.
>> For that experience of like you said discovery, you don’t necessarily start with Google
You go to Google to find an answer for something.
>> The question is real time search more navigation, becoming a navigation.
>> Potentially, but to Steve Gillmor’s point.
I would like to talk to the technical for a moment.
Anyone who is not a Facebook or Twitter at the table we’re reliant on something that’s going
On at a network that we don’t control.
And as Luic mentioned to Chris we don’t have access to Facebook, for instance, and we don’t
Have access to lots of things we’d like to.
For that developer community to grow and to continue, both Twitter and Facebook are going to
Have to do a lot of listening.
I don’t envy the position.
Seems like a difficult one to need to listen to developers like me hammer on you to say, hey,
Can you add this.
>> What’s your wish list?
>> Social graph would be my number one.
I would like to be
I think it’s awful that.
>> What’s that mean?
>> The friends list.
And like being able to filter my friends from Facebook like we can do it now with list and
like I think it’s the problem of rebuilding your friends list in both Twitter and
Facebook is kind of painful for our users.
We can see that.
And then filter them.
Why would I fill them once on Twitter for a list and vice versa.
>> Seems pretty reasonable.
>> Good. When do I get it?
>> I don’t think we’re actively working on it.
But we should be.
So I’ll get back to you.
It seems very reasonable.
Most of these things, as you all know, developing products yourselves, come down to sort of
More things to do than people to do them.
But if something becomes like a big demand and gets on our radar.
>> Ron, in your portfolio, do you
>> Just a second.
If people have questions come up to the mic.
We’ll call on you.
>> Ron, in your portfolio, real time companies, what do you think are the number
, two and three issues that third party developers are dealing with?
>> Well, I think a lot of them are trying to realize and define what is the best use case
For these products.
And how they fit into the ecosystem.
>> That’s one.
>> You have consumers who don’t really realize how to use all these services, and where all
These services fit in.
And therein lies the massive opportunity for start-up innovation.
But it’s usability.
It’s filtering, which is where we started this.
Filtering all this massive amount of information.
And the interesting thing is many companies exist today we’re an investor in several sentiment
We’re an investor in four real time search engines.
We’re investors in real time
>> Is that the list of all your investments right there?
>> It’s a smattering of real time.
There’s behavioral targeting companies.
The point I’m going to make is the technology exists independently.
>> That’s investigative journalism right there.
>> Thank God I don’t bring anything confidential to these conferences anymore.
Because last June, at this conference
there was something confidential.
So I don’t bring
>> Some of these haven’t been announced.
>> So what I was going to say, you have all these independent technologies that have been
And the innovation is going to be now that all
now all these technologies are going to be
Integrated together to give the user an even better experience.
>> I feel Mike’s leg shaking.
>> You’ve got geo location.
Companies are going to take these independent technologies and integrate them to make a fabulous
Experience for the consumer.
>> Semantics geo location, and.
>> Behavioral targeting.
>> And intent.
>> I think Mike has a question for you.
>> Just when you write stealth next to the company, is that
>> Does that mean
>> Don’t read it on stage.
>> That means I shouldn’t read it on stage.
>> I’ll give this back.
Lili, you basically Ray Ozzie, brought you on top of an umbrella of research projects.
Outlook actually rubber meeting road types of things.
From your perspective, as sort of the captain of the Microsoft real time effort, what are the
Top three priorities that you see in terms of the issues that we’ve been discussing.
>> I think one thing to remember is you have to keep the people who are inputting the information
Motivated and interested to continue to do that in an ongoing way.
So maybe what we’re interested in from a search perspective is really different than what the
Teenager just tweeting to their friends or who happen to take a picture of some event cares
So I think it’s really important to remember both the consumption side, from the real time
Also it’s just a communication tool for most people.
And the fact that we can kind of get this feed and use it for other purposes is really a byproduct
But we shouldn’t forget that for most people it’s actually a way to talk to people that they
So I just think it’s
it’s important to focus on both sides.
>> And the impact on the company?
The impact on Microsoft.
I alluded to the disruption that’s going to occur around Office as it moves to a service.
>> If you think of enterprise in general, so many enterprises are around the wall.
Around the enterprise.
And containing the information.
And so much information actually exists outside the wall.
So I think the question is like how do you manipulate that boundary?
How do people feel safe because people aren’t always comfortable with whatever paper is in
Front of them get dragged off around in public.
There’s a private boundary people have, but how do you be more open.
Like what’s the right openness you want.
Obviously it’s really easy to bring all the information in so you can seam it all together
But what goes out?
>> Questions from the audience?
>> Question: I’m bartoss Slovic.
I’m curious, a lot of what you’re talking about involves current human psychology, the way
any of the companies take a broader stance in the greater time line of history, take a look
At societal problems.
Take a look at sort of global trends that have occurred in time, I mean all things being equal,
Human beings are still human beings.
We’ve evolved slightly throughout the centuries.
But here we are now with these tremendous tools and communicating with one another in real
Are any of your companies or organizations actually facilitating like I guess steps towards
Addressing some of the global trends that have occurred throughout the ages?
>> Well, I just want to make a general comment.
You said that society is slowly changing.
I think society is radically changing behavior which is created the market that most of our
On Facebook, even in their private pages, are willing to disclose so much.
So the evolution that’s happening is posting on Twitter, that it really is public.
It’s been alluded a lot here.
Kids on Facebook, when Facebook photos started, didn’t realize that those pictures would be
Public to all their friends.
So it’s not an evolution.
It’s a radical, radical change that we need to help the consumers cope with.
>> I think having that discussion that Ron alludes to is part of thinking about what the societal
Impact can be.
I don’t know if you’re asking if one of our sites or businesses are trying to get behind a
But the fact that we’re platforms to allow people to communicate and share ideas is a major
Upside, but also a downside, as you point out.
And I think the fact that some of us think prudently is because we’re worried about the implications.
And I think it’s not that our users are stupid.
I think it’s they sometimes, you know, in the thick of things, in using all sorts of new functionalities
And technologies that come out of a daily basis they actually don’t think of the implications
What can happen to themselves.
>> I guess what I was alluding to, as you reveal information in real time, you can influence
Your decisions in real time and you can effectively close the loop on how we choose to create
Our reality in real time.
>> I’ll look at it from a slightly less lofty position.
But certainly as I have been looking at real time.
I mean, and I look at it from a programmatic nature, our social network also has broadcast
Mechanisms and areas where people come to be programmed, too.
And the old method, the portal method, was you put something up on a page.
You leave it there for a couple of days.
But we’re going to start to make programming decisions in real time so that the information
That you’re seeing around a television show and movie or piece of music or a social cause is
Hopefully more likely what you’re interested in.
And that is going to be an evolution so that just like a social network, your experience may
Be very different than somebody else’s experience on a social network.
Your programming experience should also not be a one-to-many but also with very specific one
That is driven by real time decisions.
>> Want to address that then we’ll go to questions.
>> I’ve observed in my lifetime as an engineer on search.
People used to wait a few hours to get information sometimes a few days.
I was a search engineer back then.
There was not that much of an Internet.
As time progressed and the creation of evolution happened on the Internet.
People expect to get information right away.
In any of the search engines.
So we made people’s thinking much faster.
Human beings are more efficient today than
however, the search engines a few years back
Still gave you information that was layered.
And what people are going to expect in the coming year is getting information which was produced
Seconds back to you when it’s relevant to you.
And society is
it’s going to change again.
I can’t predict how.
But I’m going to tell you my kids are going to demand more of my search engine than I do of
My own search engine because they will say, oh, come on how can this not be found by me two
Seconds after it was produced.
So societies will change.
Human beings will become even more efficient, and that’s the fun part of being in this business
Is that our great intelligent users collectively make the world better by developing interests,
Doing what they want to do and the time to that has been short circuited.
That’s the fun of real time.
>> Question: First thing I want to say is thanks to the people for being here.
This is a lively discussion.
Thanks for that.
I think the big trend over the past five years, I don’t think anybody will argue, is what’s
really happened is the number of people able to publish the Web is really expanded, whether
It be in context of a blog, social network, Twitter, real time I sort of is I guess the recent
Step in that direction.
So my question for you all, is before we talk about the problem of how do I filter a stream,
How do I index a stream, I want an assumption here that what is published to the Web, we’re
Satisfied with that.
A different way to ask the question is that there’s a big difference in the quantity and quality
Of what I know versus what I’m saying.
And what you’re publishing to the Web is this much.
And what we’re talking about is indexing and organizing and getting access to this much data.
And so I’m curious, is what we’re all talking about today a step towards accessing and sharing
What we actually know?
Or does it filter it?
>> He wishes he could think something and it will appear on the Web.
>> What you’re alluding to is a very, very reasonable point with Web pages moving on to blogs.
Moving on to tweets, it’s been easier and easier for human beings to create information, communicate
What they know.
However, that is still a bottleneck, and you’re really picking on the bottleneck that experts
Ron is not telling all about his investments, what we should invest in.
So experts who know, you know, experts who know and they would like to share still the amount
Of effort it takes to get all your knowledge available in the form to the world is more and
I think many companies will be working in that direction, and I think as that cycle becomes
Easier for people to [indiscernible] to produce.
I think from Web pages to blog posts to tweets we’re moving in the right direction.
But there’s still a bottleneck.
>> It’s knowing what question to ask.
You’re asking the question through the prism of what we’ve done in the past.
So the on boarding of a tremendous amount of signal from location applications is not about
You writing something.
And it’s immensely important for the future of solving this problem.
And there will be other forms like that which are not about someone typing in a thought or
A piece of knowledge.
It’s about our action in the world.
That’s what the real time Web is about.
It’s about the consciousness.
It’s what we’re doing
>> It’s not necessarily knowledge, though.
Not true knowledge as well.
So I worked at blogger.Com back starting in 2000.
And what we saw when we would talk to companies about blogging, even in enterprises or small
Businesses, they would say that sounds a lot like knowledge management.
And it’s very simple.
All you need to do is sign up all your employees and get them to put all their content in everything
They have in their head in this form categorize it.
See whether or not their manager can see it.
If the rest of the company can see it.
They talk to their director, and we were like that sounds terrible.
Why not just use a blog.
It’s a little white text area.
You type in there and some stuff happens.
So I think as we’ve made the text box smaller, it’s become even easier to express that knowledge.
So I think I’d have to agree with Amit that there’s
it’s becoming easier for that to happen.
I don’t know that some sort of like brain probe is going to help us like get true knowledge.
But the easier transmission, the, making the creation tool is easier is going to help make
That repository larger. I think it’s difficult to get inside someone’s head as we’re talking
But one part of the product that we all have a lot more control over is surfacing the content
That is meaningful.
And doing that in a way, if you look at Facebook and Twitter and Friendfeed, the real lightweight
Feedback features whether it’s like or retweet or comment, if those are lightweight and 46less,
I think we end up with a ecosystem that’s able to bubble up when something comes out of the
Mouth that’s valuable.
And that as builders of these mediums, we really need to focus on that.
It’s something we have more direct control over than how do we get you to say something that
>> Go to the next question right here.
My name is Adam Bolt with Goso.
I’m currently working with most of your APIs.
And today one of the things that I wanted to ask you, I hear you speaking a lot about privacy.
And be resolved at a user level.
For example, Facebook just changed a promotion guidelines.
How do you see all of the large social networks aligning themselves in that respect?
>> Sort of a standardization of rules?
Just because even like recently with Facebook’s new update, you know, with the stream and everything,
You know, what if somebody Twitters on the same page and someone does a promotion on at which
Twitter that goes over
that gets published on Facebook, it just seems like it’s a big challenge
>> Would you like to address that?
To say right now, but we’re having tons of discussions about that.
And as these platforms become more and more open, I expect that will be the trajectory.
I doubt they’ll all be like the same terms.
But I think simplification is strategically the right direction.
For everyone on stage.
Because any barrier you put in place to develop or building on your platform is bad.
I’m not sure specifically what
I’m happy after this to talk to you about specific
In general, our platform, the terms we put in place are about preventing automated spam mechanisms,
Not something that a user does.
A user has control over their own stream.
Whether or not it comes into a client or not.
>> I was talking specifically about the promotional guidelines you just announced.
>> Probably I imagine unless it’s interesting, I can chat with you afterwards.
>> Let me ask you, you probably won’t be able to answer this, but what are the terms of usage
Of Twitter data?
>> I really don’t know, as I said.
>> Do you happen to have any idea whether you can retain that data for any length of time?
>> I don’t know the details.
>> I actually don’t know the details either.
So I’m sure I can find
>> I’m sure, people who
>> So one more question over here.
>> Question: I’m Kelly.
I was wondering, do you think that search will become sort of obsolete in the sense that we’ll
Get results as users produced based on what we actively do?
And if this is the case, I know it’s a sharing of information based on your network of friends
Versus public information.
So won’t this information need to be integrated in terms of producing those search results?
>> Isn’t Google’s host on that?
>> Search already uses a lot of context.
Google’s search if you have signed on is free and uses a lot of context based on what you’ve
Done in the past.
Our social search experiment to enhance your search experience, you’re asking a very good question.
What we observe out there, there’s so much information being produced, that I would say search
Uldn’t say search will become obsolete but it has to incorporate itself to incorporate much
More information that would be available to search engines.
So the need to find relevant information would still be there because there’s so much happening
You don’t want to know about everything all the time.
There’s specific things that you need to know now.
Or 10 minutes from now.
So search would incorporate all the contextual and personal signals and search would become
Much more personalized.
With things like that my friends talk about this I need to know about this.
If I’m interested in it.
Much more geo specific search context, much more social specific search context.
And much more real time search context.
Coming together to provide search of tomorrow which are we are all working towards in our business.
>> We have about
one thing we haven’t touched upon here.
How are we going to make money off the stream?
>> We’re going to talk about this later on in the day.
But I want to get some ideas up here.
Ron, I think last time you threw out a number that the real time industry is going to be worth
$5 billion within a few years.
I don’t know whether you were talking about evaluation or revenues.
But what you see monetization to start to be turned on.
Obviously you can talk about search, you know, search advertising, but even searchtizing traditional
Search advertising don’t work with real world search because it’s hard to predict what people
Are going to be searching, there’s problems with that.
There’s in-stream promotions and advertising.
Give me a sense of where do you see the most promising areas from an economic point of view.
>> It was very encouraging this morning to hear Dick Costolo say that in 2010 sometime that
I’m looking at the list I cited.
That I cited last June.
You know, the acquisition of followers.
Is going to be worth something.
People will have to figure out how to monetize that.
>> Those are all great.
I think there will be a lot of money to be made.
But search there are just gobs of money to be made.
The reason is because you get their intent.
You get their intent at a time they want something.
And I think Erick said the biggest problem with real time search is how do you predict the
Not from the user’s perspective.
You probably get the information back to the user.
But how do you predict for the advertiser?
Sarah Palin came out yesterday with all this stuff, God bless her.
She probably created probably 100 million search queries around her picking Glen Beck as a
And what were the ads, there were no ads.
It was unbelievable.
I think the biggest ad on Google was to go try the search on Bing, and I mean
I mean that’s
>> I think your example is an outlier.
>> I don’t think so at all.
Think about every day.
I haven’t looked at the news today.
I’m sure there’s a ton of searches going on today
>> That doesn’t go to intent.
That’s the problem with news model.
It doesn’t map to intent.
Unless you want to buy her book.
>> But there’s probably a lot of ads that would want to be solved.
>> Models around geo location.
I don’t know if this is what you meant when you said incentives, the idea if you always have
An on mobile device you are walking down the street.
There’s some sort of deal.
I worry about implications and push notifications so your phone is vibrating when you’re walking
Down the street here.
I think that’s there’s lots of opportunities for that.
>> That’s the future.
Search is not going to die it’s going to be good for what it’s good at.
In the word we’re going to be using is Sam bee yent streams, that’s around us without us taking
Any action that are relevant to what we’re doing that’s going to evolve on Facebook that’s
Why it’s critical that’s why they don’t want you going elsewhere.
>> And discovery.
Discovery is a huge, huge opportunity.
Once the IP is perfected for intent of the user, there’s a whole bunch of technology to still
Be developed here.
This market is early, early days.
And I will absolutely stand by my statement that in five years this is a multi-billion dollar
It will have monetized billions of dollars.
Because this technology is being developed right now.
Google and Bing would not have done a deal with Twitter, if they didn’t think that data was
Not monetizable and very, very valuable.
>> I agree with Ron.
I think what’s important to note in being early days on anything, look at YouTube.
Google still experimenting with the ad trials there.
They don’t know whether or not it’s going to stay a banner rolled over the ad or if it’s going
To be pre-roll, post roll or if the text ads monetize better than the other stuff.
I think for us to make a presumption right now on the best monetization strategy for
>> We’re going to do that at the end of the day.
>> I think one of the things
>> Make sure Dick Costolo is around.
>> I alluded to earlier we’re talking about these sort of aggregators and portals and search
Engines that aggregate this real time data.
One thing the note is when social networks sort of took off, say five years ago or whatever,
It was now.
One of the biggest impacts on the Internet was every product you use daily became social.
Whether it was Flickr, which applied social context to photos.
You know, with Last FM and applied sort of, Pandora, applied social context to music.
One thing, I think one of the biggest impacts of real time becoming a trend is all the verticals
Of products we use will incorporate real time discovery, personalization.
And I think that like I imagine five years from now looking back for me as a user that will
Be one of the biggest impacts.
It becomes an expectation of all users in all product categories.
I think it will have a great deal of impact, it will be disruptive to all those industries,
Just like social versions of these products were disrupted to the industries in the past.
And so that’s when I was alluding to Facebook Connect before I think it’s something that we’re
All working in this space like all looking at a single fire hose.
I think as a start-up, I would be thinking, if I applied real time discovery, real time personal
Ization to this category, how disruptive could it be.
I think that’s where a lot of interesting business models fall out.
>> That’s going to be the last word.
>> Can I launch a
to the people who are watching if you can hash tech the CrunchUp and
Let’s come up with the right vocabulary for how to call the things around the bend if it’s
Let’s start having the words to talk about these things.
I think that’s part of our challenge is in these last two conferences.
Is we need to develop the vocabulary of how we refer to this so we don’t stumble over ourselves
With semantics as we try to construct ideas and have discussions.
If it’s not just search, what is this notion of information and it’s not just push and my phone
Vibrating in my pocket.
What are the terms, what do you suggest?
>> This is all good fodder for the rest of the conference.
Please give a round of applause to our panelists.
And before we break, we’re going to have a 15-minute break.
we did something interesting where there were so many companies that wanted to demo
That we decided we were going to let people opt in, attendees, and I’m going to pick two names
Right now from this jar.
And whoever name I pick come up here and you will get to demo later on during the demo panels
Before and after lunch.
So the first, oh, Julian from Superfeeder, are you in
are you ready to demo?
>> Yes, come back stage during the break.
>> Come back stage.
You’ve got about an hour to figure it out.
And who is our next lucky winner.
Somebody from bobduku.
Congratulations, are you ready to demo?
Well, come back here and we’ll explain to you the whole situation.
It’s going to be a little bit rough.
But we’re big believers in audience participation.
We’re going to break for 15 minutes.
There’s, what, refreshments out there.
Do we need to make any other announcements?
>> There’s one announcement, which is
is the mic off?
>> What’s the announcement?
>> Luic wants to give somebody in the audience a ticket to Le Webb.
>> How do we do this?
>> It has to be somebody who wants to go, that can get themselves to Paris.
>> We’re going to give away a ticket to LeWeb before you go.
>> CrunchUp of the Web, explain
>> Only ask if you can really use the ticket.
>> How do you know it’s somebody from here?
>> It will be somebody watching or somebody in the audience because they’re the only ones
That know about it.
>> Tweet out CrunchUp LeWeb if you want two tickets.
>> Just explain why.
>> Explain why you want to go to LeWeb.
He’ll pick it and what’s the time period?
>> Next hour?
>> By tonight.
>> Explain in one tweet why you want to go.
>> By 8:00 p.M. Tonight.
>> 1500 Euros each.
>> That’s 1500 Euros each.