15 ms·

Paul is fundamentally wrong with one thiing here: Larry knew google was going to be huge from the beginning. His brother Carl was already a multimillionare in

by dekhn 3y ago

Paul is fundamentally wrong with one thiing here: Larry knew google was going to be huge from the beginning. His brother Carl was already a multimillionare in tech and taught him everything required to set up a future successful company. Larry made it clear to me that he always planned Google to be a large cash-generating cow to invest in long-term AGI and there were only a few times at the beginning when he truly feared that wouldn't happen.

zeroxfe 3y ago
> Paul is fundamentally wrong with one thiing here: Larry knew google was going to be huge from the beginning. Nobody can know such a thing -- especially turning around a billion-dollar business, a thing with a near-zero prior probability. It's more reasonable to say that Larry had the drive, the aptitude, and the resources to maximize his odds (which would still be low.)
- programjames 3y ago
  If you're one of the smartest people in the world, working in one of the most impactful areas... the posterior probability is more like 90% than near-zero.
  - dekhn 3y ago
    Depends- for example, duirng the time that Google started up, I was in grad school with a great deal of support for starting biotech companies, but when I looked... even the geniuses were struggling because at the time, biotech companies could take a decade or more to be profitable or fail. It was clear, as we were exiting the late 90s dot com crash that the economy and money for tech was going to come back, and companies that took advantage of rapidly increasing specs on cheap machines (versus buying expensive Suns, DECs, or SGIs) were going to be able to scale to immense amounts of traffic, and deliver ads for profit.
    - programjames 3y ago
      I don't think at that time biotech could be considered very impactful. After all, search engines were literally impacting billions of peoples' lives within a few years, but a new drug would take a decade to get to market, and impact "only" millions of people.
- jononomo 3y ago
  Do you remember the dot com boom of the late 90s, though?
shubhamjain 3y ago
Do you have any sources for this? If he knew it was going to be this huge, why were they looking for an acquisition for north of a million dollars? On AGI: are you referencing this clip? [1] To me, it seems like an ordinary vision when you stretch your imagination. He didn't have any realistic idea how Google might achieve AGI, just something that could happen in the future. And this was in 2000. Google, I imagine, was pretty successful by then. Larry Page's original ambition was to digitize books and knowledge in the world [2]. > Page had always wanted to digitize books. Way back in 1996, the student project that eventually became Google—a “crawler” that would ingest documents and rank them for relevance against a user’s query—was actually conceived as part of an effort “to develop the enabling technologies for a single, integrated and universal digital library.” [1]: https://twitter.com/jam3scampbell/status/1608270969763729415 https://twitter.com/jam3scampbell/status/1608270969763729415 [2]: https://www.theatlantic.com/technology/archive/2017/04/the-tragedy-of-google-books/523320/ https://www.theatlantic.com/technology/archive/2017/04/the-t...
- dekhn 3y ago
  my source is larry himself- I used to work there and volunteered at SciFoo, which he attended and I chatted with him about it. The digital library is basically building a corpus for training AGI. His dad was a computer scientist and larry read a lot of AGI sci fi as a kid, so it's not really that hard to extrapolate 1980s technology to 2040s outcomes.
  - actionfromafar 3y ago
    It's only in hindsight that training with a huge corpus of text is going to give AGI. (Still not obvious if you ask me.)
    - dekhn 3y ago
      I'm the same age as Larry and started working in machine learning in the early 90s, and to me, it seemed pretty obvious at the time that AGI would ultimately need a large corpus of text (and video) to be useful. What's kind of funny is the models I worked with at the time- hidden markov models of (biological) text anticipated the validity of this approach, although by design, they don't work with long contexts.
alexsereno 3y ago
You know I remember hearing this way back in the early 2000s but forgot about it, Carl is pretty low key, no Wikipedia / etc
- dekhn 3y ago
  I chatted with him several times. One thing I've noticed is that Larry and Carl both have some behaviors that are consistent with ... inflexible thinking ... that I recognize in myself and have worked to correct. Anyway ,we chatted about his latest company, which I can't find now. The idea was this: industrial facilities often return really noisy signals on the electrical supply lines, and the electrical company has to do a lot of work to clean it up (hey, I'm not an EE). Like, motors, etc, all severely distort the waveform. So his company made and I guess sold a product that you'd put at your electrical service panel that "cleaned up" the signal before it went back to the power company. In response, the company would give you cheaper power prices. I was reminded of this when Larry Page met with Donald Trump early in the Trump administration and brought up one of his favorite plans- to rebuild the US electrical grid around DC distribution. He's truly made for another world.
  - alexsereno 3y ago
    That’s a super thoughtful response, thanks. Also my armchair understanding is that HVDC electric is superior and our electrical grid needs the rework anyways haha

by dekhn 3y ago

zeroxfe 3y ago

> Paul is fundamentally wrong with one thiing here: Larry knew google was going to be huge from the beginning. Nobody can know such a thing -- especially turning around a billion-dollar business, a thing with a near-zero prior probability. It's more reasonable to say that Larry had the drive, the aptitude, and the resources to maximize his odds (which would still be low.)

programjames 3y ago
If you're one of the smartest people in the world, working in one of the most impactful areas... the posterior probability is more like 90% than near-zero.
- dekhn 3y ago
  Depends- for example, duirng the time that Google started up, I was in grad school with a great deal of support for starting biotech companies, but when I looked... even the geniuses were struggling because at the time, biotech companies could take a decade or more to be profitable or fail. It was clear, as we were exiting the late 90s dot com crash that the economy and money for tech was going to come back, and companies that took advantage of rapidly increasing specs on cheap machines (versus buying expensive Suns, DECs, or SGIs) were going to be able to scale to immense amounts of traffic, and deliver ads for profit.
  - programjames 3y ago
    I don't think at that time biotech could be considered very impactful. After all, search engines were literally impacting billions of peoples' lives within a few years, but a new drug would take a decade to get to market, and impact "only" millions of people.
jononomo 3y ago
Do you remember the dot com boom of the late 90s, though?

shubhamjain 3y ago

Do you have any sources for this? If he knew it was going to be this huge, why were they looking for an acquisition for north of a million dollars? On AGI: are you referencing this clip? [1] To me, it seems like an ordinary vision when you stretch your imagination. He didn't have any realistic idea how Google might achieve AGI, just something that could happen in the future. And this was in 2000. Google, I imagine, was pretty successful by then. Larry Page's original ambition was to digitize books and knowledge in the world [2]. > Page had always wanted to digitize books. Way back in 1996, the student project that eventually became Google—a “crawler” that would ingest documents and rank them for relevance against a user’s query—was actually conceived as part of an effort “to develop the enabling technologies for a single, integrated and universal digital library.” [1]: https://twitter.com/jam3scampbell/status/1608270969763729415 https://twitter.com/jam3scampbell/status/1608270969763729415 [2]: https://www.theatlantic.com/technology/archive/2017/04/the-tragedy-of-google-books/523320/ https://www.theatlantic.com/technology/archive/2017/04/the-t...

dekhn 3y ago
my source is larry himself- I used to work there and volunteered at SciFoo, which he attended and I chatted with him about it. The digital library is basically building a corpus for training AGI. His dad was a computer scientist and larry read a lot of AGI sci fi as a kid, so it's not really that hard to extrapolate 1980s technology to 2040s outcomes.
- actionfromafar 3y ago
  It's only in hindsight that training with a huge corpus of text is going to give AGI. (Still not obvious if you ask me.)
  - dekhn 3y ago
    I'm the same age as Larry and started working in machine learning in the early 90s, and to me, it seemed pretty obvious at the time that AGI would ultimately need a large corpus of text (and video) to be useful. What's kind of funny is the models I worked with at the time- hidden markov models of (biological) text anticipated the validity of this approach, although by design, they don't work with long contexts.

alexsereno 3y ago

You know I remember hearing this way back in the early 2000s but forgot about it, Carl is pretty low key, no Wikipedia / etc

dekhn 3y ago
I chatted with him several times. One thing I've noticed is that Larry and Carl both have some behaviors that are consistent with ... inflexible thinking ... that I recognize in myself and have worked to correct. Anyway ,we chatted about his latest company, which I can't find now. The idea was this: industrial facilities often return really noisy signals on the electrical supply lines, and the electrical company has to do a lot of work to clean it up (hey, I'm not an EE). Like, motors, etc, all severely distort the waveform. So his company made and I guess sold a product that you'd put at your electrical service panel that "cleaned up" the signal before it went back to the power company. In response, the company would give you cheaper power prices. I was reminded of this when Larry Page met with Donald Trump early in the Trump administration and brought up one of his favorite plans- to rebuild the US electrical grid around DC distribution. He's truly made for another world.
- alexsereno 3y ago
  That’s a super thoughtful response, thanks. Also my armchair understanding is that HVDC electric is superior and our electrical grid needs the rework anyways haha