Eqs

Monday, January 14, 2013

Adventures at the Googleplex (Part II)


This is the second part of the "Adventures at the Googleplex" non-blog. As long as I don't write about something that is current, I'm still not a blogger. (I'm fairly sure that is true.) Again, what follows below are unedited ramblings written down very late at night. Possibly slightly inebriated. Quite possibly.

August 5, 2007

"This is day two of SciFoo at Google. It is 12:28am and I desperately want to go to sleep. A full day exposed to this group can wear you down. Even if it is by trivia."

The day started with the self-organized sessions decided upon yesterday. I signed up to see Jeff Hawkins talk, naturally. Well, I signed him up for it, really. And when I come into the room, Jeff points to me to explain, "Well, blame Chris Adami for this".  I am happy to oblige. Jeff launches into his talk (there’s only about thirty listeners, in an auditorium that can hold 120). He is very reserved rather than brash, almost as if he expected rampant opposition by the assembled scientists. But it is clear that he knows more about neuroscience than any of the people in this audience (me included), and he moves on to his implementation of his cortical algorithm as if this was just another obstacle to cross. He is being interrupted to outline his algorithm in more detail. And as he explains it, it slowly, but surely, dawns on me. This is it. This guy actually, unbelievably, against all odds, figured out how the brain works. Now, if you know me, I mean, know me at all, you know that I don’t give praise lightly. But in this case, I’m speechless. I’ve thought about brains for twenty-five years. I applied to graduate school in 1985 to study Artificial Intelligence (and had to settle for Theoretical Physics instead). I’ve been rejected at least as many times as Jeff Hawkins has been. But he had an idea I did not have. (Well, if you read his book, the initial idea was not his either). But who is counting. As far as I’m concerned, I know how the brain works now, and Jeff Hawkins figured it out.

Can he prove it? Well, that’s a longer story. Within his new company, called Numenta, he is trying to develop the most basic implementation of his algorithm, and that is going slowly.  But, from a fundamental point of view, I have rarely experienced such a feeling of gratification. For the last 25 years I have tried to understand the algorithm behind our brain function, and only now I think I understand it, but I am livid, nay, outraged, that I did not think of it before.

Jeff Hawkins, how dare you!

Martha Stewart talks about cuisine on the ISS at SciFoo
I’m sure there were other talks before I get to the one by Martha Stewart. The one where I snapped her picture talking about space cuisine.

Neal Stephenson is next. We sit next to the Google cafeteria/coffee contraptions, where you can have anything you want, and you don’t have to pay for it. Neal is shy. We talk about writing (as if I had written anything of consequence). And talk more about writing. Neal, in the end, is just one of the guys invited to talk to more guys (it’s mostly guys here, I have to say).

It is far too late for me to be able to paint an accurate picture of what it is like to be at SciFoo. I decided to focus on just a few of the people here, and wish, in the aftermath of it, that I would have spent more time with some of the other people I recognized: Freeman Dyson and his son George and daughter Esther, Eric Drexler (that I only talked to during an awkward lunch period when I didn’t know who he was). And finally PZ Myers. We sat on a sofa and chatted. He’s a blogger. The one everyone knows. Titus Brown wants me to say hi to him. 


Post scriptum (January 2013). George Dyson had a photo essay about  Sci Foo 2007, for those who want to reminisce. I can barely remember it, and really only because I wrote these notes. 

Saturday, January 12, 2013

Adventures at the Googleplex (Part I)

I'm not a blogger. I'm fairly sure I'm not. And just to prove to myself that I'm not, I will start off this blog with a blog I wrote from another time. When life was different, when not everybody was blogging about something. It is 2007. I was invited to Google, along with a number of people from all walks of life, to a thing called "SciFoo". On day one, this is what I wrote. There is no editing. This is what I wrote then, into a MS Word document, which I'm sure nobody ever read. The point is, just a few people were blogging then.  I think.

August 4th, 2007


"If I wanted to, I could touch her hair. Martha Stewart is sitting almost right in front of me. To my left is Jeff Hawkins, founder of Palm and Handspring, and I’ve been talking to him for the last half hour. As Tim O’ Reilly, of, well, O’ Reilly, reminded us just after singing “I Feel Good”, is that this isn’t your average scientific conference."

The bus taking us back from the Googleplex to the Wild Palms Hotel just pulled in. It is 12:23 am, but there is really no point in trying to go to sleep: my room is right next to the patio of the Hotel, and everybody is gathered there to discuss and digest the evening. So, instead, I do what probably ten percent of the SciFoo attendees are doing: I blog.

This is, as anyone can imagine, a fairly interesting experience. We were picked up from the Hotel at around 5:30 pm, and arrived at the Googleplex 30 minutes later. As we pulled in, a strange vehicular contraption, piloted by about fifteen men and women, pedaling in an indescribable way (as they were arranged in a circle around the contraption) crossed the path of the bus. None of us had ever seen anything like that, and some speculated that this was arranged. But, instead, I think the Googleplex is perhaps more like the television series “Eureka”. We check in. As at every conference around the world, I get a badge with a string. But unlike at other conferences, I also get a piece of paper on which I am to describe what I do in five words or sentence fragments, and nominate five people that I think should be at next year’s installment. Among the five, I nominate Al Gore, and a certain developmental biologist who knows Python. Then they take your picture, and then you get the SWAG. I’ve never gotten anything interesting at conferences before. (Well, when I ran Alife 6, we had California sunglasses). But here, I get:

A t-shirt
An MP3 player with a podcast of a tour of the Googleplex
A DNA puzzle
A piece of crystal (heavy!) with one of four themes engraved in them. I choose the “Kepler spheres”, since I have the same already as a 3D metal prototype by Bathsheba Grossman. I could have selected a map of the star systems of the universe. As one of the Google staffers put it: if you are abducted by Aliens, this is your map home. Google  Maps or Google Earth or Google Mars ain’t gonna help you there.
A “Google”-engraved leather notebook. Of four colors, I chose black.

Then it’s off to dinner, buffet style. I sit at the table with Paul Ginsparg (who came up with lanl.gov, or arxiv.org, as it is called now), Steve Benner, who makes enough money on royalties that he founded his own institute, and who I’ve known for the last three years, and Chris Wiggins of Columbia University, who I’m roping in to help us with modularity measures. He promises any support he can give. During dinner I’m looking around if I can spot anyone I know. I see Pam Silver, Erik Winfree, say hi to Drew Endy, and go back to discussing Martin Nowak and the Templeton Foundation with Benner and the Opinions Editor of Nature Magazine. She is constantly quizzing me about Geoff West from the Santa Fe Institute. I don’t know why.

After dinner, we congregate to do two things: introduce ourselves, and then determine how we are going to spend the next two days by putting talks into the slots of a giant matrix we are to fill with colored markers. That’s the idea of SciFoo: we set the agenda, we decide who gives talks. There are about 6 different rooms available, each with a one hour timeslot. But the rooms have different capacity. You can put down your name for a talk/discussion between, say, 9:30-10:30am for a room holding 5 people, or 150 people. Depends on the size of your ego, I suppose.

During introductions, we are to say who we are, and say THREE WORDS about what describes our passion. Because we are 250 people. I’m thinking: how many are able to do that? I describe myself as: Evolution, information theory, and quantum physics, not all at the same time. This turn of phrase becomes quite popular, when people describe their three phrases, but insist, instead “at the same time”. A number of people are incapable of restraining their self-descriptions. A gong is instituted to cut them off. It doesn’t matter how famous you are.

I want to talk more to Jeff Hawkins (remember, he’s sitting just to my left). Frankly, I can’t believe my luck. When the time comes for people to sign up for talks, I just tell Jeff that I’m going to sign him up. It helps that I was reading his book on the plane (Thanks Dimitris!) Jeff is a little timid. I encourage him: “Go sign up!”-- “You think I should?” –“I’ll sign you up if you don’t!”.

I boldly go to the big white board that the masses have congregated around, grab a marker, and write “Jeff Hawkins: How to make intelligent machines” into the 9:30 slot of a 120 people room. I figure that he can fill it. After that, I stand in front of the boards for a long time, wondering whether I should talk about something. I look at what other people are suggesting. In the end, I take a slot for a room of 15 people (that is the size of most rooms, but it goes down to 5), to talk about using evolution for design and discovery. I don’t have a prepared talk or anything, I’ll just open my mouth and begin. If you know me, you know that I can do that.

After this exercise is finished, there are a number of short planned talks. Drew Endy is the first, and he gives a version of a talk I heard him give several times. Always fun, but not new to me. After an interesting talk about the basics of energy household on the Earth (we consume 18 Terawatts of energy per day, but tidal energy only produces 3.7, so forget about trying to use that, etc. Also, energy from waves: just 3 Terawatts. Forget it.) Charles Simonyi, who used to work on Microsoft Office, talks about how he spent his summer vacation.  On the ISS. You know, the International Space Station. It’s called “Charles in Space”. The man is genuinely moved about his experience; there is a long Q&A period after, where he talks about what this was like, how it used to cost 25 million dollars, but now the price tag is more like 35 million. And how Martha Stewart prepared the banquet they had in space. Really, I’m not kidding. When it comes to space food, he defers to her. (They are dating, don’t you know?) She wanted to design the whole menu, but found out that NASA takes a long time to approve new foods for the Space station, and she had to grudgingly accept that French chef AlainDucasse already had designed foods that met NASA approval, and was reduced to prepare a banquet based on those. Oh, and Charles Simonyi doesn’t like fish.

After this talk, we walk to the main hall, where a bunch of tents have been erected, and the bar is set up. You have to understand. The original O’Reilly Foo Camp was a real camp. People sleep in tents. And Foo, here, stands for “Friends of O’Reilly”.

They wanted to have a bar, just so that they could have a foobar.

If you understand that, well, welcome!

I walk back with Jeff Hawkins. He tells me that he wants to get back so that he can prepare the talk I signed him up for. Instead, I talk to him for almost another hour. We talk about his Redwood Neuroscience Research Center, and about Numenta, his new company that implements the neuroscience he learned. I talk about my project evolving controllers for robots. Basically, we talk about trying to get something novel done when the rest of the world doesn’t think you can do it. This is great. I can sense the same passion in him that drives him to understand, but he is a totally different type of human being. He is at home in the world of technology ventures, and realizes that progress is so much faster in a company as opposed to academia. Yet, we both are driven by the same motives. Jeff tells me that he has given up control of the Redwood Neuroscience Research Center, and focuses fully on managing Numenta, that implements the ideas of his research. Anyone want to buy stock in that company? Not once did I think about the fact that this guy is probably worth a couple of hundreds of millions. ‘Cause he’s just another curious misunderstood guy to me.

After midnight, the last bus leaves the Googleplex, and I am stuck sitting next to an “inventor” on the bus back to the Hotel. I’m explaining evolution to her. How we use evolution to come up with things the human brain cannot come up with. But I don’t think she understands. She yawns a lot and slurs her speech. I, on the other hand, am alert, but wish I had a better seat mate.

But now it is 1:25 am, and we start again early this morning. Who knows what that day will hold. But I’m optimistic. Sergey Brin is there with his wife and his father. He’s got what, 4 billions? I need to talk to him. Part 2 is here.

[Note: links added recently]