Stephen Wolfram Notes

20120202-001301.jpg

Stephen Wolfram
“Science, Mathematica, and the Making of Wolfram|Alpha”
2012-02-12
4:15pm

Wolfram Recording

“People just create crap data.”

“It at first seems a little crazy.”

“This seemed very interesting thing to do, but it’s very hard.”

“Every ten years or so I re-asked, is it now possible to do this?”

“If you have an operation you want to solve, automate as much of that operation as you can.”

“If everything can be a symbolic expression, it let’s you have a very coherent language.”

“We spent incredible amounts of effort trying to do this.”

“I could have hooked it up to make it do something, that would only take a few seconds.”

“Now we can use this image as a piece of symbolic data.”

this is crazy, he really knows this language

“implement every well defined algorithm.”

“How do you model things that happen in the world?”

“sometimes this approach of finding math behind things doesn’t work.”

“What kind of primitives should you use for these rules?”

“If the rule is simple, the hypothesis might be that the output must also be simple, but the question is, what’s actually true.”

“This is quite a bizarre thing.” > rule 30 of cellular automata

“What’s the secret that nature uses to make all this complicated stuff?”

“computational irreducibility”

“We assume we are smarter than the things we are studying.”

“We are computationally equivalent to the systems we are studying.”

“If you try to define life abstractly you run into problems.”

Life is defined historically, the same with intelligence.

“What if our physical universe is a rule that exists on a higher universe?”

“That’s a separate topic.”

“There is an alternative approach.”

“You look at them and say, I don’t know how this works.”

“Things that don’t look like they would ever be created by a human.”

“Find the universal constructor.”

Can cellular automata be used for compression?

“There is nothing that special about intelligence, it is just computation at some level.”

“The next big step in computational knowledge is coming soon!”

“This internet connection is really slow.”

“The real thing I want to do is not this.”

“That is surprising.”

“It’s always interesting to me how much quantitative stuff there is to figure out in the world.”

“What do you mean by that?”

“Almost none of the data is foraged from the web.”

“Just having the raw data is only a very small part of the problem.”

“To actually answer questions you have to make the raw data computable.”

“We actually have to have experts in the loop if we want to get the right answer.”

“The whole of Wolfram|Alpha is about 15 million lines of Mathematica code.”

“We decided to just try and do it, and we figured out a lot of new things.”

“People use a kind of psuedolanguage, and I think it’s actually pretty close to people’s underlying thought processes.”

“Wecause we have so much infromation about the world, we can disambiguate with very high probability.”

“What do you actually produce as output?”

“notable people”

on the web, 10 billion interesting pages, 50 billion less interesting pages

“When can we answer questions by computing the answer rather than by looking it up?”

“The fraction of queries that we understand is getting bigger.”

“The whole thing is actually very complicated and I’m often surprised how well it works.”

“I’ve been yakking for ages here.”

“It’s actually interesting.”

“It’s interesting how you can combine free form linguistics with a programing language.”

“This is just a random blob of structured data.”

“Being able to take data from the outside world and being able to take data that one uploads and being able to merge those.”

“Pretty dangerous to press that button.”

“It’s just too easy to use this.”

“The simplist thing is to wrap the data around like a cylinder.”

“Our universe is not as complicated as it could be.”

“It has to all emerge from the low-level rule.”

“When you’re at that low of a level, there are quite a number of formulations.”

“Bascially any time you can crack what it’s doing, it’s not our universe.”

“It’s a big technology adventure.”

“I have numbering schemes but they’re not quite as easy as one-dimensional automata.”

“We’ve found most people want the whole report.”

:siri: most famous person born in springfield il

“We think we have a pretty good scheme for doing that.”

“I think we have a good, human-understandable way of doing that.”

“There are all kinds of questions I made differently because of the ability to compute the answer.”

computable document format (CDF)

“You mentioned that intelligence is computable, do you think it is possible to use the rules to figure out some ideas?”

“We know what peope want.”

“There are lumps of computational data.”

“We can make it really really fast.”

20120202-120447.jpg

20120202-120507.jpg

20120202-120518.jpg

20120202-120644.jpg