7 ms·

> Larry suggested we go way beyond this. His idea was to ‘just OCR’ the entire computer screen continuously so that you could easily search and find anything yo

> Larry suggested we go way beyond this. His idea was to ‘just OCR’ the entire computer screen continuously so that you could easily search and find anything you had ever seen on your computer. This guy founded Google? I wish I could have seen the horrified look on the engineers' faces when they heard this idea.

nostrademons 13y ago
Larry is legendary for absolutely crazy ideas. One of his other ones was to remove all the buttons from the Google home page: http://skrubu.net/2009/10/07/the-day-google-was-blank/ http://skrubu.net/2009/10/07/the-day-google-was-blank/ I think this goes to show that the secret to having better ideas is to have more ideas, and nobody cares about the million that suck, they care about the one that makes you a million.
- lifeisstillgood 13y ago
  but you cannot work on a million ideas - you have to pick the few that are actually good. Having a million ideas clearly makes it more likely you have the Big One, but I am interested in the mental process in picking and in committing. or has Mr Page done a ridiculous amount of crazy that will never work things?
  - rhizome 13y ago
    I am interested in the mental process in picking and in committing. In order to pick, you must have things to pick between. I'm not sure you can even practice picking without having a backlog.
    - lifeisstillgood 13y ago
      most of us have explicit or implicit backlogs. I think there is however something about just starting a project that tells you if it is feasible, problematic or interesting that somehow looking at a list does not. Would be interested in how Page chooses the ones he will invest personal energy into
  - nostrademons 13y ago
    He's done his share of crazy will-never-work things too. I think a lot of the reason it works out is the existence of an external feedback loop, a group of people that are willing to say "Larry, you're insane." Larry is actually pretty good at listening to other people - even now, in the New Google that sometimes does very unpopular things, he's committed to at least hearing out objections and weighing them. When Eric was CEO he also provided a very useful countervailing force to Larry & Sergey's more idiotic ideas, while still letting crazy ones like "Let's build a webmail client!" or "Let's build a browser!" or "Let's buy a mobile phone OS!" go through.
  - davidgerard 13y ago
    I suspect his results are better than those of most people here.
- ArbitraryLimits 13y ago
  Also writing "Google" on the moon with a laser.
seiji 13y ago
It's not that crazy of an idea. He's not talking about holding up a scanner to your screen, scanning it, then OCR'ing a big tiff. He's probably talking about something that takes screen shots every few seconds, OCR's that, then indexes the contents based on what's on your screen, your current URLs, etc. "Computer, search for every time I saw the words 'crazy idea' in the past six months and include how my workspace looked at each occurrence."
- gojomo 13y ago
  I'd go beyond "not that crazy" to say this idea is inevitable once compute power and storage become cheap enough.
  - LordIllidan 13y ago
    What about the surely obvious privacy implications? A lot of people on HN are barely trusting Google with their mail, or even their searches - would you trust them with seeing everything on your screen?
    - gojomo 13y ago
      Google's successful offerings in search, email, docs, desktop search, Android, and Chromebook suggest that the privacy worries of "a lot of people on HN" aren't very relevant to their bottom-line. Also, the functionality – complete screen archiving and indexing – could be implemented locally. So, its value can be evaluated separately from the question of whether you'd want the screengrabs/indexes living at Google.
    - nl 13y ago
      What about the surely obvious privacy implications What about them? would you trust them with seeing everything on your screen? Hell yeah, in an instant, if it meant I had a way of searching everything I'd seen! I'd want a way to turn it off at times ("Incognito mode"), but to me the benefits would be huge.
    - dkuntz2 13y ago
nl 13y ago
It's the opposite of a stupid idea. The one problem with it is that it isn't ambitious enough - it only is helpful with remembering stuff you saw on your computer screen. If only there was a way to OCR everything you saw, index it all and allow you to search for it again in the future. Imagine if I could wear some glasses or something, and then they had an app that could do OCR or everything I saw! That would accomplish the "OCR the screen" thing, but expand the scope to things off the computer screen too! Oh. Wait. http://www.google.com/glass/start/ http://www.google.com/glass/start/ Damn Larry Page.
- 001sky 13y ago
  If only I had a map at 1:1 scale...! Never get lost =s

nostrademons 13y ago

Larry is legendary for absolutely crazy ideas. One of his other ones was to remove all the buttons from the Google home page: http://skrubu.net/2009/10/07/the-day-google-was-blank/ http://skrubu.net/2009/10/07/the-day-google-was-blank/ I think this goes to show that the secret to having better ideas is to have more ideas, and nobody cares about the million that suck, they care about the one that makes you a million.

lifeisstillgood 13y ago
but you cannot work on a million ideas - you have to pick the few that are actually good. Having a million ideas clearly makes it more likely you have the Big One, but I am interested in the mental process in picking and in committing. or has Mr Page done a ridiculous amount of crazy that will never work things?
- rhizome 13y ago
  I am interested in the mental process in picking and in committing. In order to pick, you must have things to pick between. I'm not sure you can even practice picking without having a backlog.
  - lifeisstillgood 13y ago
    most of us have explicit or implicit backlogs. I think there is however something about just starting a project that tells you if it is feasible, problematic or interesting that somehow looking at a list does not. Would be interested in how Page chooses the ones he will invest personal energy into
- nostrademons 13y ago
  He's done his share of crazy will-never-work things too. I think a lot of the reason it works out is the existence of an external feedback loop, a group of people that are willing to say "Larry, you're insane." Larry is actually pretty good at listening to other people - even now, in the New Google that sometimes does very unpopular things, he's committed to at least hearing out objections and weighing them. When Eric was CEO he also provided a very useful countervailing force to Larry & Sergey's more idiotic ideas, while still letting crazy ones like "Let's build a webmail client!" or "Let's build a browser!" or "Let's buy a mobile phone OS!" go through.
- davidgerard 13y ago
  I suspect his results are better than those of most people here.
ArbitraryLimits 13y ago
Also writing "Google" on the moon with a laser.

seiji 13y ago

It's not that crazy of an idea. He's not talking about holding up a scanner to your screen, scanning it, then OCR'ing a big tiff. He's probably talking about something that takes screen shots every few seconds, OCR's that, then indexes the contents based on what's on your screen, your current URLs, etc. "Computer, search for every time I saw the words 'crazy idea' in the past six months and include how my workspace looked at each occurrence."

gojomo 13y ago
I'd go beyond "not that crazy" to say this idea is inevitable once compute power and storage become cheap enough.
- LordIllidan 13y ago
  What about the surely obvious privacy implications? A lot of people on HN are barely trusting Google with their mail, or even their searches - would you trust them with seeing everything on your screen?
  - gojomo 13y ago
    Google's successful offerings in search, email, docs, desktop search, Android, and Chromebook suggest that the privacy worries of "a lot of people on HN" aren't very relevant to their bottom-line. Also, the functionality – complete screen archiving and indexing – could be implemented locally. So, its value can be evaluated separately from the question of whether you'd want the screengrabs/indexes living at Google.
  - nl 13y ago
    What about the surely obvious privacy implications What about them? would you trust them with seeing everything on your screen? Hell yeah, in an instant, if it meant I had a way of searching everything I'd seen! I'd want a way to turn it off at times ("Incognito mode"), but to me the benefits would be huge.
  - dkuntz2 13y ago

nl 13y ago

It's the opposite of a stupid idea. The one problem with it is that it isn't ambitious enough - it only is helpful with remembering stuff you saw on your computer screen. If only there was a way to OCR everything you saw, index it all and allow you to search for it again in the future. Imagine if I could wear some glasses or something, and then they had an app that could do OCR or everything I saw! That would accomplish the "OCR the screen" thing, but expand the scope to things off the computer screen too! Oh. Wait. http://www.google.com/glass/start/ http://www.google.com/glass/start/ Damn Larry Page.

001sky 13y ago
If only I had a map at 1:1 scale...! Never get lost =s