Previously published to my SweatInvestor.com blog. Moving it to this one as I revamp it…
(Note: SweatInvestors are welcomed to make posts from interesting events, especially if the events are “invite only”. Both to share their access as well as to share their unique perspective on the event if they have one.)
I’m sitting at a Google launch event with no idea of what will be launched. I was invited simply because of Google User Group Meetup list The secrecy is borderline thrilling, or maybe its just the coffee. Mental note: don’t tell people stuff, it’ll make them want to know about it that much more. Blind launch announcements are great.
Live Notes
9:42 - People are still filling into the auditorium at the Computer History Museum…
Internet went down, so I pulled out my trusted Sprint card (thanks Dave Kozuki!), and we are back on…
As I wait I noticed someone from Techcrunch is here live blogging as well…
10:13: Marissa Mayer jumps onto the stage…
She outlined 4 major focuses of Google…
Modes: text, images, location, mobile phones, voice, etc. More modes, will drive more search.
Media: pictures, maps, location, etc…more than links. The more modes, the more searches, the more types of media, the more searches, etc.
Language: Yada…
Personalization: Yada…
Rate of Progress: This week in search…
10:18: Vic Gundotra, VP of Engineering appears on stage…
Talked about innovation and how initial launches of them usually failed to see the full scope.
Trends:
- Computing: Moore’s Law
- Connectivity: More connections from More Devices
- The cloud: access to huge resources
- Mobile devices: video, speakers, gps, camera, etc.
By themselves, these sensors on this android phone are not so special. But within context to the cloud, its much more. They becomes eyes and ears. Search by site, location and voice.
Did demo of Google voice search on phone, the crowd applauded…
Then did a voice search in Manderin, McDonalds in Beijing, the crowd applauded more…
Then a demo in Japanese…
Good demo on real time cloud translation from English to Spanish, 2010, will enable more search…
Location…wants to make an integral part of all products. In Japan, everyone always hangs on to their mobile devices, always within 1 meter. Same here. The device’s location is our location…
(Live blogging note: Learned that you shouldn’t add pictures till you are done, unless you have auto save that doesn’t refresh. Lost some data and find that I have to wait awhile each time it needs to refresh…)
Search Suggestions based on location, SF vs. Boston on two iPhones…Red Sox vs. REI when typing RE.
Making “Location” a first class object. (I’ve worked at a Location based company before, I’ve seen this coming, and its huge. Looking forward to seeing how this rolls out and how it relates to all the GIS/LOCATION/ETC companies that are starting to be introduced to the market. Its just the beginning.)
Adding “Near Me Now”….
Sees where you are, gets context, goes up to the cloud, tells you what is RELEVANT near by…
Both whats NEARBY and WHATS GOOD…
Announcing today a new Google Maps for Android…Whats Nearby is not just on the Search Page, its now within the map app…
(Just noticed that there is Draft Auto Save….nice!)
GoogleGoggles - lets you use a picture as the query. Gives eyes to the internet…
Demoed…took picture of a bottle of wine, it then searches that wine and gives you results like cost, taste, etc…
Then, to be more practical, took picture of landmark in Japan…
Images streamed to google servers, images are analyzed, detecting signatures, matching with index, best matches ranked and then sent back to the device…
Why is this product in GoogleLabs? Because we are just in the beginning. Still lots of room to improve. Works well on some images in some categories. Our goal is much wider and far reaching. At some point pointing your finger at any objects and getting info on it, like a virtual mouse for the world….
We are at the beginning of the beginning…
10:43
Started initially online indexing links…
- Ranked URLS
- Then to books, images, products, blogs, etc. (all on the results page)
- Lots of relevancy issues, which one, when, why, etc…
- There are user updates out there that are relevant…but how do you get them, determine relevance, and then get them to you…
Amit Gundotra arrives on screen to make the big announcement in search, but first, a little history…
The history of information flow:
- Kids asking grandpa, “Should I eat those beans?” “no, they are poisonous”….passed on from generation to generation when those kids became grandpa’s…
- Then the printing press…knowledge spread around the world a lot quicker…
- Servers: billions of documents available to users via search engines…(crawled the info every months, indexed it and then published monthly…then faster, now crawls every few minutes…in today’s world, that is not fast enough…the world updates info every second, twitter, blogs, etc. Info is being created at a pace that we’ve never seen before. Seconds matter).
Announcement: Google Real Time Search
Google relevance technology meets the real-time web.
Relevance, Relevance, Relevance.
So much info out there, getting you relevant info is the key. We’ve done this the past 11 years.
Ok, demo…
Bill Casey, product manager, pulls out search:
…Searches “obama”…real time feed into the results page (funny, Tiger Woods link to naked golf also appears…)
Clicking it takes you to a real time page, someone tweeted about this #googlesearch and it appeared on the screen…
New search option “latest results”…
“Options” link goes to page where you can add “updates” link…
Example: tweet that talks about line and if the vaccine for H1N1 ran out at a specific location, saved them a trip there…
Also: Google Trends page is coming out of Labs…
“Hot Topics” section: number one is Tigers Mistresses
11:03 - Talks about the huge amount of work that goes into making this
Had to develop dozens of real time technologies to make search real time…
Query hotness, probability of relevance, author quality, tweet quality, language model, topicality, Query volume fluctuation, etc…
As the amount of information has grown, billions of documents processed a day….RELEVANCE has become the critical factor of building products like this.
Info dissemination from generations, to years, to hours, to minutes….to seconds. We are proud, but not satisfied.
We at Google will not be satisfied when the speed of light is the only barrier between you and your information.
11:07 - Back to Marissa….
Facebook will be providing a live feed from public pages, and will be included in real time search. Same for MySpace. Twitter is already in there.
Marissa Concludes: last year we laid out our vision of search (the 4 modes). Now we talk about adding eyes and ears in real time. We did a lot in just one short year.
11:10 - Opened to questions, starting with one from online viewer…
Q: will Google Goggles use face recognition from the co you just purchased.
A: we made the decision not to include this (we have the ability, but decided to delay to help safe guard privacy)
Q: what is the revenue generating ability of real time search?
A: this is still young, we are experimenting with multiple business models…you’ll find huge innovations within the next year on this. There is great real time data to draw from…
Didn’t hear the question, I think it had something to do with what data sources will you draw from…
We want to get the most feeds of realtime info, however we can…
Q: What about real time search outside of english?
A: We want all languages, but will start with English, in Q1 we’ll be launching with other languages once we’ve fine tuned it all…
Q: How prevent spammers to game the real time results…
A: We are very experienced with this. We have developed counter acting responses before things happen. Real time is moving from minutes to seconds. Matt and his team, are experts in this area.
Q: what about personalization. Will it be able to filter down to people that I care about.
A: we are very excited about whats happening. We are just getting started. If you mix the key principles that Marissa talked about (e.g., geo, personalization, etc). Merging the multiple technologies, from social search, location search, geo search…that is coming soon.
(Wow, lots of big media outlets here…I’m starting out with the players…I’m wondering if I should start handing out water or towels..)
Q: do you need realtime feeds from partners? What if you don’t have a partnership with everyone…
A: we believe the more info, the better.
Guy next to me is about to ask a question (Danny, the speaker knew his name))
Q: Can you disclose the financial details of these partnershps? Or at least if there are details?
A: we can’t disclose the financial details.
Funny back and forth, as expected, no comment. Interesting implications either way, great question…after the event as the guy from Myspace, picture right, was leaving, as if rock star, people paused asking Marissa questions on the side of the stage, and everyone, including Marissa, directed their attention and questions to him: can YOU comment on any financial components to the google/myspace deal? No comment, of course. But could have been straight from a movie screen play.
Q: Any things you wouldn’t publish? Can you talk about relevance?
A: Meh…(didn’t take notes)
An online question was then read out loud, by Luke Wilson, (crowd laughs) a Luke Wilson, not the Luke Wilson….funny odd moment…
Didn’t catch question…
Q: Is this the death of Journalism and if information is power, are you the most powerful company?
A: This is a loaded question (softened, crowd laughs). Our goal is always about bringing timely info to users. We are empowering users with this. We’ve been doing this for 11 years. Its all about user empowerment. And realtime search is just the next step. Marissa adds that its about getting people the info and then off our site as quickly as possible. She said she took exception to the question about Google “having all the information”.
A few more questions….
Term Marissa brought up “Dog Fooding” stuff, using their products for real problems as they develop it…stroller recall or something…defending that real time info can be as important or more important then traditional info objects from general search…
(Observation, live blogging brings readers closer to the event…but I feel like I’m almost not hear. I just looked up at the speakers for the first time in about 20 minutes…)
Q: can you help say what is TRUE in real time search?
A: that’s really hard, many years down the road…lots of challenges and whats true is really tough (paraphrasing…)
Comment about moving to a standard API to allow people to upload real time data to Google…that’s the direction we’d go….
Q: how does page rank apply to real time?
A: its one of hundreds of things we use. We had to invent a lot to handle real time. Page Rank is really important, but we just needed to develop many new things to make real time search as relevant as we have….(Marissa adds that page rank is also about determining authority. So its somethings we are beginning to experiment with).
Q: about geo locating
A: GPS data (can wait 20 minutes) so we fall back on cell towers (with our database, based on cell tower, we can give you a quick estimate until we get the info from your phone’s gps data).
The amount of phones that use google maps gives us help with our coverage of locatoins.
Q: will google Real Time search api be big?
A: I assure you we will be looking into the details of that?
Q: when will this all go live?
A: Japanese mobile search, google maps, and a few others are live now…others will be launched in the near future. The translation device is just a technology concept demo. Q1 will be when you may start seeing it. To check it out, go to www.google.com/trends and then select “hot topics”.
Q: so you are doing like text status updates? Any other types of info?
A: Yes, we are excited. We’ll do more as we learn about the status updates, do relevancy, etc. As time passes, we’ll be expanding to many types of data. Like TwitPic. Excellent info there. Myspace is already looking how they can take some of their non-text updates into this.
Q: what more are you doing with inventory data and search?
A: Q1 more is coming out. We have some partners, etc.
Q: Ryan from Wired.com, whats the difference of the real time and google search index?
A: Google search index has been updating every minutes. Real time search is just a continuation of that technology, just faster. The new thing we added is update receiving, updating and merging (something like that). And all the info that models how info is flowing in the system.
Closed event with a thank you. Then off stage questions….
A few observations and side conversations:
Saw people asking questions while recording both voice and video from their iPod. Its interesting how horizontal the apple product line has reached.)
There were a bunch of big wigs in the reporting industry. I’d hear, “hey Danny, you need to be Mr. x from the New York times…” and then there would be an intro. Just about everyone that asked a question was from a major publication. Was definitely cool watching and taking it all in. Little did anyone know it was the first live event covered by SweatInvestor. So, Mr. x from the New York Times, if you didn’t meet me at the event today, perhaps we’ll be introduced at the next one…
Someone asked again afterwords if Google ever pays publishers for their real time datafeeds (like Facebook or news publishers). Marissa didn’t answer this directly but did state that Google doesn’t pay to index content and essentially its up to the provider to monitize the resulting traffic (i.e., Google wants to send people quickly to where they want to go, and other sites benefit from this). She did qualify a little later that they do sometimes pay for data like map data.
Someone asked about if at some point Google plans to index “private” data, not just public real time data from Facebook. Marissa said they see that as part of the long term goal (i.e., index the world’s info, along with personalization as being a core principle of theirs) but that its real far off. But they are really interested in it. I personally was interested and closely listening here because I find this as an incredibly interesting problem to crack and I’ve been incubating a possible venture on it for the last few years. Two important questions, whats Google enabling and how not to be crushed by them?
My Wrap-Up Comments
Today was a great glimpse of what its like to Live Blog with the fishes (i.e., big media). We’ll see where 2010 will take me and how many “back stage” access events I will find my way into. I look forward to any pointers from readers who are knowledgeable of the press circuit.
Google no doubt is an amazing company with a lot of smart folks. Their focus on search is incredibly smart, especially because it clearly has become the most important component to any application. Or rather, in my opinion, is 80% or more of an application if done right.
Answering the question of whats important to a user right now, right where they are (i.e., here), continues to become more more valuable as the number of object types, contributing devices and real time updates grows and is indexed by Google…done right, navigation controls are less important, and the user is simple provided with a search results page (maybe with special presentation or refinement controls).
On a personal note…
One of my goals for 2010 is to set the foundation of getting press passes to unique events to meet and interview innovative people. So, I’m giving Live Blogging its first go. I’m wiredless up and ready to update live. If there is anyone out there that is knowledgeable about obtaining press passes, or has advice on how to make live blogging, please comment below.
{ 0 comments… add one now }