Also worked at Google in the 00s, and have spoken personally with several people who worked on Google when it was still a Stanford grad project.

To hear Larry tell it, his original idea for Google was to download the whole web and throw away everything but the links, and he was inspired by academic notions of citation indexes. It wasn’t specifically (or really at all, until Google books came out) about books - that sounds like a cover story to get grad student funding for his actual idea, because you can’t really go to a university and say “I want to download the whole web and throw away everything but the links” without telling them why or connecting it to some plausibly useful idea.

The acquisition offer is often cited, but I don’t think Larry and Sergey were ever really serious about taking it. They didn’t, after all. But if you’re living on a grad student stipend (which was like $20k in the 90s), a million bucks is a lot of money.

And they absolutely knew it would be big, they just didn’t know how it would be big. It was quite popular within the Stanford lab as well…it’s interesting, looking through the early CVS commits for Google, how much work was done by people who ended up not actually becoming employees of Google or having a major part in its story. I think the real genius of Google was treating the web as an object that could be studied and manipulated, and not as some amorphous thing you were part of. That was exciting even if people didn’t know how it would be.

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.

His dad wasn't "just" a computer scientist, he was a professor for computer science and artificial intelligence at Michigan State. Sergey Brin's father was equally a professor in mathematics at the University of Maryland.

> It's obvious in retrospect that this was a great idea for a startup. It wasn't obvious at the time.

It may not have been obvious, but the decision was damn well near as engineered with data as it could be. They also were _insanely_ well connected through their and their parents academic career. Both Sergey and Larry had obtained their PhD prior to starting the company. I can also remember reading that they obtained significant amounts of funding through connections Larry's dad had into the industry.

You can ignore the rest of my comment, what follows is just my take.

Their success story is imo one of the most blatant examples on how privilege really does give you a boost in life. I am not arguing that anyone could have done it, but I do wonder how the world would look like if we were all kids of academics with a successful career, with a relatively safe, secure and stable childhood home and a family background that really incentivizes learning and academic success over succumbing to the pressure of, you know, having an income at some point.

I don't think larry or sergey got their phd - larry got a masters and I think sergey left before he got any degree.

Don't forget that Larry's older brother Carl went through the whole VC process with egroups and gained extreme experience with how to maximize his position in negotiation.

But yes, I agree completely that kids of academics raised in an environment that incentivzes learning is a reliable way to transform the future.

The timing is a little muddled here - Scott Hassan (who, interestingly enough, wrote most of the original code for Google) founded eGroups with Carl Page in 1997, after working on BackRub/Google. The two should be viewed as parallel projects - Google actually started first, but eGroups raised money first. It’s true that Larry got valuable experience through having his brother raise capital first, but much of that also came from having supportive angels like (Sun founder) Andy Bechtolsteim, (Stanford professor and Granite Systems founder) David Cheriton, (Junglee and Netscape exec) Ram Shriram, and (Amazon founder) Jeff Bezos.

So to correct my original statement: Both have a masters degree and both were pursuing PhD's before they focused on Google. But I consider them more than halfway there, afaik both have published papers separate from the paper that eventually lead to Google. This is taken from their own testimonials from their paper [0]

The rest of my statement is true.

[0] http://infolab.stanford.edu/pub/papers/google.pdf

This is what all these blogs don't really mention and I have seen this over and over and over again. You need a fundamental place in society: really high up to attempt to do any of this.

Well you can still be Larry Page's dad (or his grandad) and attempt to give your children a launching place to become Google. Most success have to be built multi generationally.

> Their success story is imo one of the most blatant examples on how privilege really does give you a boost in life. I am not arguing that anyone could have done it, but I do wonder how the world would look like if we were all kids of academics with a successful career, with a relatively safe, secure and stable childhood home and a family background that really incentivizes learning and academic success over succumbing to the pressure of, you know, having an income at some point.

I recently read something that played out a thought experiment about "imagine the world could only hold 10,000 people". It may have been a comment somewhere on HN honestly. The idea was that if the world could only hold 10,000 people, there never would have been enough division of basic duties for any one individual to dive so deep that they could invent semi-conductors (insert any modern tech really). I mean we likely wouldn't even have mastered locomotion by now if that was the case.

Lets assume we're "all kids of academics with a successful career, with a relatively safe, secure and stable childhood home and a family background that really incentivizes learning and academic success over succumbing to the pressure of, you know, having an income at some point." Most of us will still just be working on the basic societal problems of producing food and taking out the trash.

It was on HN, this comment: https://news.ycombinator.com/item?id=39746742

> I recently read something that played out a thought experiment

With all due respect, the world isn't a thought experiment and the dynamics of a system with 10.000 people is not comparable to that of 7 billion.

My argument was pointing at the fact that it may be pointless to compare your own plans to success against someone elses success story, when they had completely different starting variables in life and people who refuse to at least consider this are either trapped in the Silicon Valley hustle culture or are, sorry, completely ignorant, bordering on idiotic.

With that same logic ala "Not everyone can be X" I can justify almost all of human suffering in the world. I refuse to believe that you follow an ideology like this if you're on this site, because if you did, you were already rich by means of some criminal enterprise exploiting humans far beyond what tech startups are capable of.

This isn't some appeal to idk, introducing communism and making everyone equal but this argument is as dumb as making a healthy runner compete in the paralympics.

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.)

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.

In the “Measure What Matters” book the author talks about investing in Google in early 1999, and the founders were indeed projecting billions of dollars in revenue, using ads.