Wolfram Alpha's response on "Who is Euler?"
Vladimir Bondarenko
http://www.cybertester.com/Private/gs2.85/Wolfram_Alpha_Euler.pdf
Google responds to the same question with the top link as follows
http://www.gap-system.org/~history/Mathematicians/Euler.html
Peter
Dear Vladimir,
please do not assume that your readers are so dumb
not to be able to make their own tests.
Asking "Who is Bondarenko?" Alpha responded
Wolfram|Alpha isn't sure what to do with your input.
* Did you mean: Who is mansakonko?
To the same question Google responded with this top link
knowledgerush.com/kr/encyclopedia/Bondarenko_crater/
So Google might (rightly?) think you are the man in the moon.
However I find Alpha's answer more creative.
Seemingly you are very interested in Alpha. So it would
be nice to read your thoughts in a more explicit form.
For example it is not clear to me whether you noticed that
the answers are categorical different: Google points to
an entry in a sorted list whereas Alpha tries to find his
own answer.
In some way this reminds me of the days computer chess
started. I bought such a machine and, believe it or not,
I did win. I was very pleased. Then I read this article
in Scientific America which told me that I have not be
afraid to lose against a computer in my lifetime.
I was even more pleased. I do not have to tell you
the rest of the story. The chess engine on my PC
has an estimated ELO rating of about 3000 whereas
I could not increase my abilities to the same level.
So what role are you going to play? Do you want to
tell us: "Look how easily a bright chap like me
can make jokes out of the dumb Alpha machine.
Do not believe that semantic information mining
will ever bring new insights on your desktop." ?
Richard Fateman
http://www.wired.com/epicenter/2009/05/blog_epicenter_0511_wolframlevy/
re Peter's view that saying Alpha is a failure is like saying computers
can't play chess (ever) in 1968.
My assumption is that Alpha will get better. How much better is
impossible to predict.
It is possible to assess the contrast between the claims of Wolfram
regarding Alpha today, and its performance today. On that basis it
seems that the claims are overblown.
Wolfram claims (see article above) that Alpha includes Wikipedia, the US
Census, and "about 9/10 of what you'd see on the main shelves of a
reference library."
Unfortunately, without a better simulation of a reference librarian, it
is hardly equivalent to 9/10 of a reference library. It is of course
possible to have much less information and do more calculation with it,
and Alpha seems to have that possibility since it has Mathematica
available, but as I've noted, the Google calculator could be extended.
"What are the ingredients in a graham cracker crust?"
"where does stephen wolfram live?"
"who died on april 15, 1865"
(Abraham Lincoln is the answer it did not find)
"how deep is the grand canyon"
none of these were answered, or even understood by Alpha
"how high is niagara falls?"
[gave the elevation of the city, not the waterfall]
finally,
"what is the height of the campanile in Berkeley, california?"
note that Google gives, for {berkeley campanile height} as the first hit:
Campanile - Wikipedia, the free encyclopedia
Sather Tower, "The Campanile" and carillon of the University of
California, Berkeley, is 307 ft (93.6 m) tall and has 61 bells. ..."
Maybe we should wait for Wolfram Beta.
RJF
Bart Goddard
$as4....@nlpi069.nbdc.sbc.com:
> Maybe we should wait for Wolfram Beta.
I typed in my name, and it came back with the distance
from Goddard, Kansas to Bart, France. 4951 miles or
9 hours. It seems to have missed the obvious: The cost
of riding Bay Area Rapid Transit to the Space Center in
Texas.
B.
--
Cheerfully resisting change since 1959.
Nasser Abbasi
"Bart Goddard" <godd...@netscape.net> wrote in message
>
> I typed in my name, and it came back with the distance
> from Goddard, Kansas to Bart, France. 4951 miles or
> 9 hours.
I read that W|A does not contain personal info.
So, unless you are really famous person, WA would not know anything about
you.
Only very important people are stored in WA databases. Once you become a
very important person (really important) then WA would answer personal
information about you, I am sure.
--Nasser
Peter
> I typed in my name, and it came back with the distance
> from Goddard, Kansas to Bart, France. 4951 miles or
> 9 hours.
;-)) You are not the only one.
Florian Rheinstetten entered his name. Here is the answer:
http://www55.wolframalpha.com/input/?i=Florian+Rheinstetten
I like these examples. Aren't informations put in an unexpected
perspective the germ of creativity, of innovation? Any yes,
this is what I expect /in the long run/ from a system like Alpha.
P. S. Interestingly Alpha gave a different answer when I asked:
Bart to Goddard | 6525 km (kilometers) | 7 hours 20 minutes
An American engine which knows the metric system!
This by itself means innovation ;-))
Bart Goddard
news:0GeRl.17236$%54....@nlpi070.nbdc.sbc.com:
> So, unless you are really famous person, WA would not know anything
> about you.
>
> Only very important people are stored in WA databases.
OK, now you sound like my mother-in-law.
Richard Fateman
...
>
> So, unless you are really famous person, WA would not know anything about
> you.
Abraham Lincoln works
Ulysses Grant does not
>
> Only very important people are stored in WA databases.
Once you become a
> very important person (really important) then WA would answer personal
> information about you, I am sure.
I think Grant has lost his chance to become important. He won the Civil
War, and was US president.
Try the classic "who is buried in Grant's tomb?" then try ask.com
for the answer.
RJf
Bart Goddard
958b-05e...@k8g2000yqn.googlegroups.com:
As an American, I'm not too familiar with this godless, communard
metric system, so could you explain how the conversion from 9 hours
to 7 hours 20 minutes works?
And I suppose someone should point out that it's going to take
at least 9 hours to get from Goddard Kansas to the seat of an
airplane that can cross the Atlantic.
Peter
>> Peter <Peter.Lusc...@googlemail.com> wrote:
>> P. S. Interestingly Alpha gave a different answer when I asked:
>> Bart to Goddard | 6525 km (kilometers) | 7 hours 20 minutes
>> An American engine which knows the metric system!
>> This by itself means innovation ;-))
> As an American, I'm not too familiar with this godless, communard
> metric system, so could you explain how the conversion from 9 hours
> to 7 hours 20 minutes works?
Well, that's easy. If you use the metric system you drive a Porsche
but when you use the United States customary system (also called
the Imperial system because standardized by the British Empire)
you drive a Buick or a Saturn.
Phil Carmody
> Vladimir and others, can see Wolfram's own words re business model.
> http://www.wired.com/epicenter/2009/05/blog_epicenter_0511_wolframlevy/
>
> re Peter's view that saying Alpha is a failure is like saying
> computers can't play chess (ever) in 1968.
>
> My assumption is that Alpha will get better. How much better is
> impossible to predict.
Given that it's entirely unable to give any useful information
at all to people using lynx, links, or w3m as a browser, and
presumably most screenreaders, I'd say it was worse than, erm,
Christ knows - WAIS? That's a step back how many decades?
Phil
--
Marijuana is indeed a dangerous drug.
It causes governments to wage war against their own people.
-- Dave Seaman (sci.math, 19 Mar 2009)
Peter
> Peter's view that saying Alpha is a failure is like saying computers
> can't play chess (ever) in 1968.
Yes. And that we will have to wait for some decades before
the /idea/ behind Alpha will become reality like those
incredible chess engines in use today.
> My assumption is that Alpha will get better. How much better is
> impossible to predict.
Alpha as a product - who knows. But the idea behind Alpha
- semantic information mining - will establish itself.
> It is possible to assess the contrast between the claims of Wolfram
> regarding Alpha today, and its performance today. On that basis it
> seems that the claims are overblown.
Sure. But that's quite typical for this industry. (For those who build
these systems and sometimes detectable even in those who criticize
them.)
> "what is the height of the campanile in Berkeley, california?"
Ok. Now it is time for an successful example. Type in
pi_15(S^8)
The answer is not so bad, I think.
> Maybe we should wait for Wolfram Beta.
"The next Google will be Google," someone said who knows this
industry.
Google is hard working on the same technology (minus NKS, of course).
Bart Goddard
04a318...@x5g2000yqk.googlegroups.com:
> Well, that's easy. If you use the metric system you drive a Porsche
> but when you use the United States customary system (also called
> the Imperial system because standardized by the British Empire)
> you drive a Buick or a Saturn.
Chevy, actually. But it used to be Schwinn, until none of
my wreches fit any of the bolts.
Tim Smith
Richard Fateman <fat...@cs.berkeley.edu> wrote:
>
> "What are the ingredients in a graham cracker crust?"
I don't think they have recipes, but if you just type "graham cracker"
you get a lot of interesting information about graham crackers,
including little graphs of the nutrient distribution and that the
serving density is 0.36 g/cm^3.
> "where does stephen wolfram live?"
> "who died on april 15, 1865"
> (Abraham Lincoln is the answer it did not find)
Try just "april 15, 1865", and Lincoln's death will be noted as one of
the notable events on that day. It usually seems to work best if you
don't ask for a very specific item, but instead ask for the more general
data, and then pick out your specific thing from the result.
> "how deep is the grand canyon"
>
> none of these were answered, or even understood by Alpha
>
> "how high is niagara falls?"
> [gave the elevation of the city, not the waterfall]
Click on the place at the top of the results where it says "use as a
waterfall instead".
--
--Tim Smith
Peter
> "what is the height of the campanile in Berkeley, california?"
> note that Google gives, for {berkeley campanile height} as the first hit:
> Campanile - Wikipedia, the free encyclopedia
> Sather Tower, "The Campanile" and carillon of the University of
> California, Berkeley, is 307 ft (93.6 m) tall and has 61 bells. ..."
> Maybe we should wait for Wolfram Beta.
Or for Bing? Six days to go.
http://www.bing.com/
And I will try again
pi_15(S^8)
I bet you win.
Jean-Michel Collard
The 15th homotopy group of the eight dimensional sphere? :)
Do it by hand !
Peter
>> Peter wrote:
> > Or for Bing? Six days to go.
> > http://www.bing.com/
> > And I will try again
> > pi_15(S^8)
> The 15th homotopy group of the eight dimensional sphere? :)
Yes :-)
Here we go:
http://www.luschny.de/temp/bingversusgoogle.png
Microsoft's Bing offers a 'eye mask' from a 'snore-shop'.
Google offers: Alpha! Ha! And indeed a link to the topic.