Why is it so hard for people to understand that this tool IS NOT an aggregator or a meta search.
:smack:
Why is it so hard for people to understand that this tool IS NOT an aggregator or a meta search.
:smack:
I typed in ‘esperanto’, and it said, “Total number of speakers 2000.” That is clearly wrong. A list of the people who registered for the 2006 world congress of Esperanto has 2183 people. And I know there are more, because I speak it and I didn’t go there!
Seems like an almanac basically. Not a very good one though. And they put in a few canned answers to standard joke questions.
I entered 1066 and it knew it was the date of the battle of Hastings.
1492, 1453, 1776 - it had no idea.
So it does math too?
Meh.
“Now establishing standard orbit around Wolfram Alpha, Captain.”
Every time I see the name.
Hmm, that’s interesting. By default it doesn’t evaluate “1066” as a date, but as a historical event.
So, it gives information such as what date that event occurred on and who was involved.
If you enter “July 4 1776”, it interprets it as a date, and gives you information about the time of sunrise and the phase of the moon, as well as notable events, in particular “America declares independence”.
If you enter “declaration of independence”, it interprets that as a historical event, and gives you the same categories of information as if you had entered “1066”
On the suggestion of a friend, I asked Wolfram|Alpha how much wood a woodchuck would chuck if a woodchuck could chuck wood.
It has to learn, right now it is Hal learning Daisy. We use Mathematica at the supercomputing institute where I work, and I’m feeling pretty confident Wolfram will push this further.
The problem about it is people are comparing it to Google, and it isn’t. The types of questions it will handle and ultimately learn are mathematically based. Its about crunching numbers.
When I played around a bit, it was hit or miss, but it did seem to do a good job of converting queries into mathematica-speak. Often times the problem was just data missing in the wolfram server. Even if it may not be great now, could something like this have been conceivable 5 or 10 years ago? If this is a sign of things to come, I’m looking forward to it.
Because it is one? And more, of course. Or supposed to be. ‘Computable knowledge engine’, they call it. To compute knowledge, you must first gather it. For instance if you plug in ‘car sales 2008’ it will tell you how many cars were sold in 2008 - in the USA. Plug in ‘UK car sales 2008’ and it fails. It doesn’t have the background data yet for the UK.
It’s a piece of shit and I would have been embarrassed to release this crap to the public.
Before opening it to the public, I saw this introductory video and was wowed.
If it ever gets to that level, for any sort of query of the type shown in that video, that will be an amazing product.
As it is, it is very limited. I don’t even know why they released it so soon if the databases are far from complete and the parser is shit.
Some real examples I tried:
-> price of milk / average salary
Wolfram|Alpha isn’t sure what to do with your input.
OK, let’s make it a bit easier:
-> historical prices for milk
Wolfram|Alpha isn’t sure what to do with your input.
Let’s make even easier:
-> milk prices
Wolfram|Alpha isn’t sure what to do with your input.
Really? You don’t know what to do with the input? Why not just say “I don’t have any information on milk prices at the moment” ?
Something new:
-> tallest man
height of the tallest person ever recorded: Robert Wadlow
8.93 feet
Yeah, something works. What about if we narrow it down a bit?
-> tallest man alive
Wolfram|Alpha isn’t sure what to do with your input.
OK, what about:
-> tallest man today
Wolfram|Alpha isn’t sure what to do with your input.
On to something new:
-> price of a pepsi
PepsiCo | closing price
$ 51.17 (US dollars) (Monday, May 18, 2009)
I actually wanted the price of a can of pepsi, not the stock price. Let’s try:
-> price of a can of pepsi
Wolfram|Alpha isn’t sure what to do with your input.
Piece of shit.
I’ve been trying some basic physics queries and I’m definitely disappointed. For instance, if I ask “drag,” it just gives me a dictionary definition, with no link to any actual physical information about drag. OK. I ask about “drag formula.” No luck. “Drag equation.” Nope. “Drag coefficient,” however, brings up a calculator and the equation I was looking for. On the other hand, it really seems to know nothing of rolling resistance. Or mechanical friction.
If you ask about “sound speed,” it interprets the query as “speed of sound at sea level and 0 deg C.” OK, so I try “speed of sound at 1000m and 20C,” and it ties to multiply the speed of sound at sea level by 1000m! After many tries, I finally got something of an answer by asking: “speed of sound in air at 20 degrees C.” If you don’t specify “air,” it calculates the result for water. In neither case does it give you the equation. The speed of sound actually changes with humidity, and I tried to make it take humidity levels in considerations. I tried. And I gave up.
I think it’s impressive once you realize that it does specific tasks and not much else. For example I tried one of the suggestions: typing the names of two companies like IBM and Apple and it immediately gives you a bunch of information about the financial performance and stock prices of the two companies. No other search engine that I know of gives this to you immediately; you would have to go into Google Finance and do a customized search to get it.
I have already installed it in my Firefox search box and I will definitely be using this quite a lot.
To echo others: the database seems a bit paltry right now. I tried:
muzzle velocity 5.56
and all sorts of variants, and it “did not compute”.