Tumblr 101: Tips And Tricks To Get Started
WATCH: Shredding On The Google Guitar!
Wolfram Alpha is live and interesting, but not a Google killer
Finally, after all the sneak-peeks and numerous blog posts about how Wolfram Alpha could finally be the Google killer prophesized in cave paintings on ancient server room walls, the computational knowledge engine is live.
Yes, Wolfram does have some snappy answers to questions you know you've been wanting to ask it, like the windspeed velocity of swallows, 42, and so on. Give Alpha something to sink its teeth into like "high temperatures (city)" and you'll see what sets it apart from the pack.
Toss it a famous personality - say, Bruce Campbell - and you'll get a very basic bio: name, age, date and place of birth. Give it a location, and you're presented with its population, current time, weather conditions, and location on a map (unless you're in the middle of nowhere like me). H1N1? Yes, Wolfram will give you up-to-date numbers on that as well.
Certain queries, of course, turn up nothing. Detroit Red Wings? Nope. Stanley Cup? Try again. Just because such terms are trending elsewhere doesn't mean Wolfram is too concerned with them - yet.
No, it is not going to kill Google. It's not supposed to. Wolfram Alpha will, however, be an immensely useful tool for academic and research applications - and a profitable one at that.
If you plan on testing it out, make sure you're wiling to wait. The Wolfram servers are being bombarded with requests, so you may encouter the occasional error message.
Comments
9
Subscribe to commentsradman1331May 16th 2009 5:23PM
google does some things similiar. i type in weather for a zipcode and it shows me it.
sitrucMay 16th 2009 6:14PM
Google killers already exist in ask, Live, Yahoo!, etc. Wolfram Alpha type results used to exist in a few different searches, but I haven't seen them much recently. It's nice, but not useful for general web searchs.
QuikboyMay 16th 2009 11:23PM
Live Search also does weather & zip code look up. As well as calorie content ("calories soda"), dictionary definitions ("love define"), embedded traffic maps ("traffic houston"), phone number look-ups, etc.; other stuff that Google doesn't do. Just wanted to point that out if you're the kind of search shortcut person.
AlexMay 16th 2009 11:55PM
The Red Wings may not work, but the Minnesota Twins (or even, "The Twins") works. I guess Wolfram has no love for Canada/Hockey
f055May 17th 2009 5:28AM
I agree with this article completely. Wolfram|Alpha is a really different beast than others and in no way directly competing with Google. It's a different approach and because of that it's the future - they showed that copying Google isn't the only way to do search and question answering. You can get my full opinion at http://f055.net/article/wolframalpha-launching-today-and-what-it-means-for-google-and-search/
wrs589May 17th 2009 7:30AM
Maybe it's just me, but they could really do with a new server?
JamusMay 18th 2009 10:12AM
Do a search for the word "iPhone". You get nada. Zip. Zero. With such a prevalent term, you would at least expect something.
Now do a search for Pluto. BOOM. You get tons of stuff related to planetoid orbits, mass, speeds, etc..
But not even one lowly hit for "iPhone"? Odd.
RosaMay 22nd 2009 2:43PM
now there is Kumo : http://www.newsy.com/videos/searching_how_to_search
Though I don't think either alternative will cut into Google's share, at most they will just spur innovations that Google will add to its system.
Your momJun 3rd 2009 6:00AM
"Wolfram Alpha will, however, be an immensely useful tool for academic and research applications - and a profitable one at that."
I don't think it will actually be useful for academic nor research purposes. Everything you can do with Wolfram Alpha is limited to a single-line text input. Any *real* academic or research problem, or any interesting problem ever invented requires more than one single calculation. You can't set your results into a variable and use it again, you can't write a function to perform the same calculation with differing inputs, you can't write loops to refine your values, and you can't do anything repeatably and unambiguously, as everything is parsed through a bunch of ad-hoc rules that are trying to guess what English sentences mean.
Even their research is bad. For most physical constants, you get fewer significant digits than is actually known, (like for the gravitational constant, which only gives you a lousy 3 digits,) so you really can't trust numbers. Anyone doing real research will go elsewhere to look up the *real* values, and calculate with a platform that doesn't truncate results to 4 decimal places. I can't imagine a situation where a researcher would be comfortable with their dumbed-down, truncated results, and not want to go to primary sources.
So you're saying they're going to be profitable? How? Explain their revenue model.
1. Curate trillions of pieces of data
2. Buy some of the most powerful supercomputers in the world
3. Incur huge expenses in maintaining data
4. ????
5. Profit!
Stephen Wolfram has said that he's not going to sell. Selling the thing while the hype is hot is the only way that he could apparently recoup even initial costs.
In short, it's not even as capable as the most trivial Turing machine.