Slashdot Mirror
Wolfram Promises Computing That Answers Questions
An anonymous reader writes "Computer scientist Stephen Wolfram feels that he has put together at least the initial version of a computer that actually answers factual questions, a la Star Trek's ship computers. His version will be found on their Web-based application, Wolfram Alpha. What does this mean? Well, instead of returning links to pages that may (or may not) contain the answer to your questions, Wolfram will respond with the actual answer. Just imagine typing in 'How many bones are in the human body?' and getting the answer." Right now, though the search entry field is in place, Alpha is not yet generally available -- only "to a few select individuals."
I don't think this can be examined without language issues. Lojban attempts to make a parsable constructed language (currently undergoing a few grammar issues, but mostly locked down). As we get closer to the Singularity, with regards to infant-style general AI and perhaps even transhuman implants (thought detector or such), we'll see perhaps a myriad of unambiguous languages.
Q: How many bones are in the human body
A: Did you mean cumulatively or at any point in time?
Sig Battery depleted. Reverting to safe mode.
It goes further than that. Try Googling "how old is Britney Spears" and "what is the population of iceland" (without quotes). The answer appears at the top, separately from the search results.
That seems to be hardcoded though, it already fails at "how old is Steve Jobs".
Trying to find mathematics/physics information is often pretty terrible. I mean, if you are just looking for a topic you can generally pull up related papers, but that is about the depth of complexity you are capable of searching for.
Unfortunately there is no convenient (or universal) plaintext notation. If you are doing anything serious you probably use latex markup (e.g., \Psi^{*}\Psi) or something similar to render images of your equations. That's well and good for people who just want to read your paper, but for people who want to do a complex search to find very specific bits of contextual information, it is just about useless.
So if I can hope that Wolfram's goal is to make his company's math and science knowledge base searchable by some sort of contextual framework, then that could be pretty awesome for those of us who would like to penetrate particular aspects of independent fields without having to become experts on the fields first.
When things get complex, multiply by the complex conjugate.
It also fails on This - seriously...
Isn't a proof of the Riemann Hypothesis necessarily non-constructive? If so, a computer can't answer your question.
After all, I am strangely colored.
True Knowledge have been doing this for over a year. Anyone can add facts to their database, and it will attempt to use those facts to infer answers to questions. Its actually very cool, although doesn't yet support such notions as uncertainty.
Try checking the number of horns on a bicorn and you'll see that the google engine is not intelligent, artificial or otherwise. Or would you like to argue that bicorns are not real, and therefore don't count?
It is dangerous to be right when the government is wrong.
It goes further than that. Try Googling "how old is Britney Spears" and "what is the population of iceland" (without quotes). The answer appears at the top, separately from the search results.
Google them together, it returns your post!
It is dangerous to be right when the government is wrong.
While this may be true for some people, it's their own fault if they limit themselves in this way. The people that are really passionate about research will use this technology as a tool to enhance their research capabilities. Those that do not probably weren't motivated enough to be successful anyway.
http://www.policystew.com/
This whole question was answered decades ago (1970s) with the "foreigner in a sealed room" turing thought experiment. It showed that the person in the sealed room doesn't have to understand english, or even know the answer to questions, provided they are given some simple rules to link words together in a response depending on what words are in the original statement.
That's not quite correct--the Chinese Room thought experiment does not depend on "simple rules"--it imagines a Turing-Test-passing program converted to book form, which is then run manually by an English speaker, responding to Chinese inputs. But there's nothing in it that implies that the Turing-Test-passing routine is simple.
In fact there's nothing that says such a routine is even possible. The Chinese Room thought experiment has always struck me as begging the question. It starts by assuming that a routine exists that has passed the Turing Test, then shows that a machine running such a routine need not demonstrate actual thought. But it is entirely possible (IMO likely) that the Turing Test cannot be passed without actual thought, which would render the first assumption void.
Build a man a fire, he's warm for one night. Set him on fire, and he's warm for the rest of his life.
That's what everyone said when wikipedia came out.
I'm a researcher, and I find that googling/wikipediaing my questions often helps me know if I'm looking in the right area. If I've conjured a sufficient amount of buzzwords about the topic, I'll get a good wikipedia page or mathworld.com page or something. I can then look at what THOSE pages reference and usually find links to some peer-reviewed sources.
Also, I've noticed a lot of my coworkers (myself included) will often try to learn more about a problem we're working on and through wikipedia/google find similar problems in several other fields. A lot of the time the other fields that have a similar problem as mine are so different than my field that I wouldn't have noticed the problem existed elsewhere without stumbling across it on the internet.
There were many editors on the book, but unfortunately, they weren't actually allowed to do anything. I mean nothing.
If an editor is denied write access to the book, is he still an editor?
I don't care if it's 90,000 hectares. That lake was not my doing.
Good points, but this is still just a different (better perhaps?) implementation of the same concept. The big issue with the implementation is that it will only "know" what you tell it, the same as any other computer.
Or human, for that matter.
The big difference is inference. If I tell you the facts "John married Jane in 1981" and "Frank is Johns son, he's 15", you will probably conclude that Jane is very likely Franks mother, at least until you get conflicting information. Computers so far could not. AI research has been working on giving them that ability for almost 20 years now. After lots and lots of failures, they've also made some progress. The big issue hasn't been the collection of facts for years now, but how to combine those facts to generate new "knowledge". That's something we humans do with so much ease that it is too easy and gets us to generate false "knowledge" all the time - marketing are experts at exploiting that, as are novel writers.
Assorted stuff I do sometimes: Lemuria.org
I would bet big money this is just another iteration of the Knowledge Scam. Read this: http://personal.lse.ac.uk/angell/papers/knowledge%20management/km.htm