Google CEO Eric Schmidt On The Future Of Search: “Connect It Straight To Your Brain”
by Michael Arrington on September 3, 2009

This is Part 2 of my series of posts summarizing a fascinating recent hour-long one on one interview with Google CEO Eric Schmidt.

Early in the interview I asked Schmidt about the future of search. I brought up the “search is 90% done” misunderstanding from last summer. Said Google Vice President Marissa Mayer at the time:

Search is a science that will develop and advance over hundreds of years. Think of it like biology and physics in the 1500s or 1600s: it’s a new science where we make big and exciting breakthroughs all the time. However, it could be a hundred years or more before we have microscopes and an understanding of the proverbial molecules and atoms of search. Just like biology and physics several hundred years ago, the biggest advances are yet to come. That’s what makes the field of Internet search so exciting.

Specifically I asked Schmidt “What are the hard things to be solved in search in the next ten years?”

His lengthy answer meandered around a central theme, that Google needs to move “from words to meaning.” In other words, Google needs to understand queries better, and return results that best match the real meaning of a query. “We have to get from the sort of casual use of asking, querying…to “what did you mean?”"

He then took a detour and shared a (non-serious) approach that cofounder Sergey Brin has talked about internally – direct brain implants:

Now, Sergey argues that the correct thing to do is to just connect it straight to your brain. In other words, you know, wire it into your head. And so we joke about this and said, we have not quite figured out what that problem looks like…But that would solve the problem. In other words, if we just – if you had the thought and we knew what you meant, we could run it and we could run it in parallel.

When I (again, jokingly) asked if Google was working on that product, he answered “Well, I wish we were. But we don’t exactly have all the medical clinics necessary to test brain insertion.”

But he also had a serious point. One big problem with search is a proper understanding of what exactly the user wants. And then how to pair that with exponential growth in datasets:

Okay. So to me, the question is sort of, what’s next, is really basically how far does the artificial intelligence technology go here? How many signals can we get from who you are, where you are, what you’ve been, what you’ve done and so forth to refine that querying? And at the same time, you also have this enormous expansion of data sets. I think what people are missing is that the amount of information on the Internet is growing very, very rapidly…Because it gets more open, people put more data on it and so forth and so on and that’s wonderful. Also, you have all these dynamic databases that are now – they basically publish that at web pages and again index them as well.

The long term goal of Google search, he says, is to give the user one exactly right answer to a query:

So I don’t know how to characterize the next 10 years except to say that we’ll get to the point – the long-term goal is to be able to give you one answer, which is exactly the right answer over time. Okay, you know, the question I’ll ask today, how many Americans have – what percentage of Americans have passports?…The Google’s answer was a site, which was somebody who had attempted to answer that question and had multiple answers. It’s quite interesting actually to read…So you go to a very good definitive site. And what I’d like to do is to get to the point where we could read his site and then summarize what it says, and answer the question…Along with the citation and so forth and so on.

More interesting topics from the interview coming up soon.

Advertisement

Responses

Comments rss icon

  • Neural implants and the techniques needed to make them work are actually plugging along in places like Janelia Farm. Seriously cool work. http://bit.ly/1TfeCR

  • Fascinating stuff, are we seeing the video at the end of these posts?

    • yeah, but the video was only the last 10 mins or so of the interview.

    • “One big problem with search is a proper understanding of what exactly the user wants”.

      this is the biggest internet search “myth” you will ever hear. for years the print yellowpages and newspapers know what you want and has everything classified into common sense categories. the future of search is the company that can best “rolodex the interenet” . search has fundamentally always been about “location”.

      search as we know it and cyberspace is for the birds, 99.999% worthless.

      as search improves it will definitely be shrinking from the spam results Googl returns today. i dont know about one answer but a handful is definitely better than the 5+ million we see today.

      the Aloqa recommendation channel startup is a sample of whats comming. brace yourself. its facinating.
      http://vator.tv...endation-engine

    • Nice to have Part2 of the interview. This makes clear that Google will empower itself until it could make a most accurate and refined search.

  • Yes, nothing new here, but glad to hear that they are thinking along these lines. Indeed, the only way to give the user the answers he is looking for while minimizing the raw size of the query is to perform the search in the context of his mind and ongoing thoughts, thus a brain interface is most likely going to be a requirement. Looking forward to it myself.

    • It’s Interesting how Google search can be fairly uninteresting because it’s so unclear how Google is improving. That must be hard for Google because I’m sure they make advances but no one notices. Twitter is clearer.

    • “One big problem with search is a proper understanding of what exactly the user wants”.

      this is the biggest internet search “myth” you will ever hear. for years the print yellowpages and newspapers know what you want and has everything classified into common sense categories. the future of search is the company that can best “rolodex the interenet” . search has fundamentally always been about “location”.

      search as we know it and cyberspace is for the birds, 99.999% worthless.

      as search improves it will definitely be shrinking from the spam results Googl returns today. i dont know about one answer but a handful is definitely better than the 5+ million we see today.

      the Aloqa recommendation channel startup is a sample of whats comming. brace yourself. http://vator.tv...endation-engine

  • Hell yea, why go to school when you could just “think it” and google could tell you instantly. Sure hope that app doesn’t go down though. hehe

  • Facebook is the Google killer!

  • They’ve actually already got the brain-search technology. They’re just waiting to launch it until they can serve ads to you as you dream.

  • then they are trying to be like wolframalpha !! but without the interference of humans,,,,,,be original!!

  • Facebook is the Google killer!

  • They’re just waiting to launch it until they can serve ads to you as you dream.

  • This is in direct contradiction to the dates they’d previously given for the launch of Google Mind. With Gmail being down the other day and now this, Google is seriously slipping.

  • Interesting insights into Google. Looking forward to more articles in the series!

  • Interesting article. Looking forward to more of the series in the future. Keep them coming!

  • people who say twitter and facebook and any of these bs social networking site is going to kill google or is next google are f**king idiots… also google is not next microsoft either… google is one of those rare companies in technology field thats trying to solve groundbreaking science problem… i would like them to see getting more into AI and even understating human mind…

  • Transhumanism on the rise. Welcome.

  • Has search engine ever thought of sorting there results by subject. Most of the time user ( that is me) knows what subject I am looking into but I will be getting results from bizarre topics that I have no interest in. The biggest problem for the search engine is to understand the page and categorize it into the specific subject, field, sub field and the topic itself. Like Entertainment -> Music ->Artist ->American ->Jazz -> …. Essentially doing what Google says it wants to do. Organize the data of the world in human understandable way.

    The idea of having to search images separately just kills me. It need to be a option to isolate those but it should not be excluded from the main search. By the way, it should extend to all the special categorize today’s search engine carry. Lets hope bing will ping the google.

  • “the long-term goal is to be able to give you one answer, which is exactly the right answer over time”

    assuming of course that all questions have *one* right answer…which i no longer believe to be the case. some do, but many don’t.

  • It seems we wont have to move any other organ apart from the brain in coming future, what we want we just need to have a thought of it and all would be fulfilled.
    Technology as they say it is in for another turn around. Let us see how long does it take to really get this technology in reality.

    Sonal Maheshwari
    USourceIT your single source for all IT needs

  • Lol, that is cool, “connect it straight to your brain”. I have heard rumors people in silicon valley are working on a computer that runs on organic hardware like your brain, maybe we need to do that first.

  • OMG Google wants to turn us into the Borg. “Resistance is Futile”.

  • Pretty exciting to see Google going this direction. This is consistent their investment in and backing of Kurzweil’s Singularity University at NASA Ames. Surprised though that Schmidt still talks about data as living on static ’sites’, and not streams and sets.

  • What OS does your brain run?

  • “where we could read his site and then summarize what it says, and answer the question…Along with the citation and so forth and so on”

    Is Google going to kill everyone, i.e. will they get into the business of returning the actual result with bit of citation (like time and weather today) or will they send people to sites.

    Google needs to be regulated and broken up into pieces as it’s stifling innovation except it’s own.

  • it’s a great vision, but google is still the category, and we need to diffrentiate many kind of searches: domain specfic, age specific, explaratory etc (some searched might well start from other organs than the brain btw).

    Problem is that search is currently an infrastructure oriented market, and only the huge companies can work out the expensive server farms.
    So maybe the solution is really open search API that will encourage startups to innovate without the impossible financial burden of thousands of servers infrastructure.

  • Arrington… you sit down with Eric Schmidt for an hour long interview, and this is what we get? A multi-part series with entirely too little focus? Come on man, leave the ‘to-be-continueds’ to the cartoons and give a piece like this the time it deserves.

  • No one’s going to stick anything in my brain if I have anything to do with it. Sure, they’ll tell you it’s very helpful and useful….at the start….but what’s next – behaviour manipulation, programmed responses, mass manipulation – I have major worries about such a brave new world – what about individual freedom?

  • Google needs to re-invent themselves which is hard to do as they need to rework their core technology.

  • If the Terms of Service of the implant is similar to the ones on the web, I will plug the implant into my cock so they can check what I think.

  • I am Locutus…

  • And then came Wolframalpha……..

  • If google will wire the brain, others will want that as well.. FB for example

  • Mike — i agree with @swkolupailo. Give this interview its due and release it. I’m interested, but the way your feeding it to us like giving me bread one day, then peanut butter the next week, and finally jam another few days hence. Getting it all at the same time let’s me make a sandwich … and by sandwich i mean context, inference, flow, tone etc. You should know better.

  • C’mon Mike — giving it to us piecemeal is like bread one week, pb the next and jam on the third. Give it to us all at once and we can make a sandwich. Like this … we are hungry. You know better.

  • Google is hard-wired to advertising — pay-per-click — that’s one problem. They’re talking about inverse search and they just won’t admit it yet. http://groups.g...ack-forum?pli=1

  • Wish you have asked how Google is losing in anonymous tests to yahoo (and even Bing)… Google in search is only a brand and not really the best results…. The first step they should do is be better than yahoo before getting into human brains…

  • Not so sure the “connect search to your brain” comment was a joke. Nicolas Carr mentions it in his book “The Big Switch” and says that the concept has always been Sergey and Larry’s ultimate goal.

  • “… And what I’d like to do is to get to the point where we could read his site and then summarize what it says, and answer the question…Along with the citation and so forth and so on.”
    Wolfram Alpha maybe ???

  • My real test case for searching:

    Type: “Is it raining outside?”
    Expected Answers: Yes/No.

    Actual: :(

  • Is this guy really the Google CEO? Is it April 1st? We expect to eventually give you 1 result? Yeah right.. that would be a horrible situation, and an invasion of privacy to ever reach that level. Plus, we like four or five results. If I was taken right to a result, even if right, I would be pissed off, not happy.

    • A result does not have to be a hyperlink to some resource. Rather, it can be a single, short report assembled for you on the fly. In the future, such a report could be built for you by combining all the interesting results into a one-pager, with a storyline generated automatically and possibly with animated graphics, and a newspaper style layout, so that you can scan it quickly. Or it could be a video, or both… your choice. Also, the sources will be authenticated so that you can indicate what kind of sources you would rather trust. In short, TRUSTED STORYTELLING will be the name of the game. But, hey, that is just my opinion.

  • 78 million people, 30% of the US population over age 14, have passports. The up-to-date number was in the fifth result returned by google, a USA Today article about record numbers of passports.

  • One problem I see, and that Google and everyone should consider, is that there is often more than one single possible answer to a given question.

    Whatever the technology used and how integrated into our brains it will become, as long as the answers will be given by private interests, as Google, I guess it will be questionable. And obviously, I don’t think we could rely on them to answer such a question!

    Google is great, don’t get me wrong. But I also appreciate the option to have my own brain!

Leave Comment

Commenting Options

Enter your personal information to the left, or sign in with your Facebook account by clicking the button below.

Alternatively, you can create an avatar that will appear whenever you leave a comment on a Gravatar-enabled blog.

Trackback URL
Short URL
bugbugbugbug
Techcrunch on Facebook