Blog anniversary

Another year has flown by here at Wolfram|Alpha, and the gears are really turning! New data and features are flowing at a rapid rate. To celebrate, Wolfram|Alpha’s creator, Stephen Wolfram, will share what we’ve been working on and take your questions in a live Q&A.

Please join us on Facebook or Wolfram|Alpha’s Livestream on Wednesday, May 18, 2011, at 10am PDT/12pm CDT/1pm EDT/6pm BST.

If you have a question you’d like to ask, please send it as a comment to this blog post or tweet to @Wolfram_Alpha and include the hashtag #WAChat. We’ll also be taking questions live on Facebook and Livestream chat during the webcast.

We’re looking forward to chatting with you on May 18!

31 Comments

W|A should be growing exponentially but isn’t. I think it is because the
initial vision is not being followed in regard to:-

- Computative result. Some output does not even include a result pod and
usually most of their output cannot be carried forward into a higher level
query. All output should be suitable for input. This will allow multilevel
query/questions that will
be mind-blowing and force powerful people to use it to compete. Any serious
thesis/idea would need to be supported by a single complex W|A query that
proved the point citing sources for all the data and algorithms used. A
contribution to a scientific journal/serious publication, would require it.
Executives would demand it of their subordinates.
.
- Curated data: Wolfram cannot collect and maintain enough data itself. It
must concentrate on curating the curators. This will allow Wolfram to divert
the resources to more valuable tasks.

- Consistency: W|A must integrate all data added. It should have revealed
some serious inconsistencies in the present data and thus generated
publicity .

- Viability: W|A must generate profits for itself and its users to finance
faster growth. The above changes and a profit model are essential.

Posted by Brian Gilbert May 15, 2011 at 6:52 am Reply

I have just received an email saying that my earlier comment should have mentioned that I am a Volunteer Curator.

Posted by Brian Gilbert May 16, 2011 at 9:38 am Reply

Hi,

Thanks for Q&A. I have two questions:

1) When will the localization of Wolfram|Alpha start?

2) What are your plans to have up to date volunteer curated data? After data are submitted, how often will you check its validity?

See you!

Posted by nanoturkiye May 16, 2011 at 1:28 pm Reply

When will the site be in Text? The output in graphics is almost useless – why so paranoid? Do you really think people will steal this stuff?

Posted by Don G May 16, 2011 at 1:58 pm Reply

Will IBM Watson Wolfram Alpha collaboration advance Natural language processing ? Is there a possibility of merger between Watson technology wolframalpha. Do you think it will be beneficial in name of science to join forces in this ultimate quest ? keeping aside corporate politics. After all both are insanely challenging research efforts and such a collaboration may enhance our lives .

Posted by Tarun Vij May 16, 2011 at 2:07 pm Reply

Hello- I use both Mathematica 8 and Wolfram Alpha on a regular basis. I am a plant scientist and agricultural entrepreneur and I have to say that agricultural production and plant related data is extremely weak. I suggested several data sources several times such as FAO or UN ComStat but they are not available. In addition, solar radiation data is not available either, which is available from NASA. The gist is that there is really limited stuff for us plant scientists and more would be appreciated, especially with the recent emphasis on sustainability and agriculture. Frankly, any sources related to these two initiatives would be most appreciated and relevant to many people. I think it might make sense to appeal to a new segment of users out there who need this type of data but are not as numerically oriented as other analytical disciplines.

Posted by baki kantasi May 16, 2011 at 2:56 pm Reply

Where do you see Wolfram Alpha in a year from now? What wonderful new features do you foresee playing an important role in everyday life?

Posted by Stephen Beckman May 16, 2011 at 3:29 pm Reply

Congratulations on your initiative that has provided new and exciting horizons of knowledge in all sciences. This innovative way to group knowledge dawned on my daughters new interest in mathematics and their contributions to the studies.

What is your guidance for the teaching of mathematics for the new generations?

Posted by Carlos Alberto May 16, 2011 at 3:37 pm Reply

The bottleneck in computability is stark and clear. As always, it’s the human cognition computable world interface. Have you ever spoken to Jeff Hawkins? It seems to me that you two could be the next Watson and Crick.

Posted by Michael Will May 16, 2011 at 8:15 pm Reply

    Thanks for the alert! I found it very interesting that Jeff considers memory as a predictive, interpreting the information in seeking to anticipate events.

    Posted by Carlos Alberto May 19, 2011 at 6:48 am Reply

Does W|A actually use cellular automata as part of the processing for queries? If so, how? What underlying automata are used, if there are some that are dominant?
Does W|A use processors that are optimized for processing cellular automata, or would it if they were available?

Posted by Ben Towne May 16, 2011 at 10:06 pm Reply

    Hi Ben,

    While cellular automata are not used directly in the underlying algorithms for Wolfram|Alpha, the methodology used to build the algorithms have aspects of the rule-based structures described in Stephen Wolfram’s book A New Kind of Science.

    Thank you!

    Posted by The Wolfram|Alpha Team May 19, 2011 at 3:14 pm Reply

W|A has been a boon to pre-college science education. Now that NYS curriculum recognizes active data interpretation and analysis as one of many skills important in science education my colleagues have been designing activities to model real world applications in technology (e.g. energy production), physics (e.g. stellar evolution, ) as well as others. Mathematica 8 is a valuable tool for teachers but is difficult to use for high school students and isn’t always easy to operate on our tightly managed equipment. Are there considerations to provide a subscription version of W|A (Wolfram Beta?) that would provide a bridge between the capabilities of W|A and Mathematica (such as persistance, simple symbol definition, etc.)?

Posted by Paul Beeken May 16, 2011 at 10:19 pm Reply

Can Wolfram be more informed by suggesting related terms?
I mean when I ask a question, It doesn’t work. then if I suggest some relatives (then people – human being can catch what i want to ask), can you develop Wolfram to that edge? Adding relatives suggestion?
Or I go back and try to make input right? it’d pretty hard then.

Posted by Hien May 16, 2011 at 10:33 pm Reply

A New Kind of Science demonstrated that many simple systems in nature were able to perform universal computation, of which Wolfram|Alpha was a consequence.

The importance of computers and computation in modern world is undisputed. However, the manufacture process of computers as of today involves rare materials, and thus might not be sustainable without absolutely strict recycling.

Since many systems in nature can perform universal computation, it’s possible to sustainably build computers with common materials, and thus preserve the computation heritage as of today.

May I ask whether Dr. Wolfram can provide some thoughts on this?

Thank you very much.

Posted by Yuting Liu May 17, 2011 at 3:32 am Reply

Most Wolfram Applications are based on data with a reliability of near 100% such as mathematical formulae. W|A is differrent with the contradictory reliability of say global warming data at say 50%. A complex query involving say 3 pieces of data each with a reliability of say 50% would have a compound reliability of .50% times 50% times 50% giving 12.5%.
W|A therefore needs to include the reliability in its computations and the result.

Posted by Brian Gilbert May 17, 2011 at 4:59 am Reply

Will this Q&A be recorded? Facebook is blocked at my school and thus I will be unable to watch it live, but I would still love to watch it when I can.

Posted by David May 17, 2011 at 8:12 am Reply

The source information is really not useful – until each data points are individually linked to source, the information generated by WAalpha is not that trustworthy. Unless WolframAlpha is trying to become a source itself, which I don’t think makes sense as it is not the source of anything but a site that aggregates data. I’d like to know where the data is coming for surely your curation process captures that. If not – then Wikipedia is a much better resource if I have to depend on crowd sourced data. You can’t expect users to just trust you or do you?

Posted by Kevin C May 17, 2011 at 3:45 pm Reply

I am a college student, taking calculus.
I really enjoy the way wolfram is able to show pictorial representation of the area between two curves.
I was wondering about the possibility of being able to solve volumes of revolved areas with the same pictorial representation.
Thanks

Posted by Justin F. May 18, 2011 at 12:19 am Reply

Do you think that [space] = 1 = [mass]^0[length]^0[time]^0[current]^0
should be a new SI unit

based on
Plancks[mass] = 2.176114e-8
Plancks[length] = 1.6160510e-35
Plancks[time] = 5.3905634e-44
Plancks[current] = 3.479+25

this gives a gravitational [force] of = 1.21056211324e+44
[mass][length][time]-2
and a [velocity] of 29979296.818358
[length][time]-1
and [energy] of 1956332455.11666
[mass][length]2[time]-2
most importantly at Plancks scale we see a gravitational acceleration of
[acceleration] = 5.56143346386314e+51
[length][time]-2

do you belive this to be the correct gravitational acceleration at a Planck scale that
might merge quantum mechanics and general relativity ?

Posted by Mark Baker May 18, 2011 at 8:00 am Reply

    This is not the right place where to ask. Please post your question on W|A forum.

    Posted by _blindsniper_ May 18, 2011 at 10:29 am Reply

I have been using W|A since the release. However I got disappointed after no seeing any evolution on it. For instance, W|A has been trying to replicate curated data, formula results and math knowledge in the traditional way.

In my opinion, a big step to W|A would be to show knowledge in an unconventional way. More intuitive, using a natural way. Fundamental math will be more attractive if you can show it using its nature. I will be father soon and I am struggling to find a way to show math (sciences) to my daughter in a more natural way, such that she has not follow the conventional wisdom. If W|A could help doing that, it will be a really milestone in science and a daily tool for new generations.

Questions for Prof. Wolfram:
What is your intuition said about this?

Do you think will be better to generate results in this way rather than try to reproduce curated data?

Trying to process data and present in a more natural way will require a more natural machines. Have you tried to build (hardware not software) machines to process data?
For instance the Wolfram’s 2,3 TM?

Posted by Elkim Roa May 18, 2011 at 10:44 am Reply

The livechat is password protected. I assume after asking a question I will receive via email the pass. How is is work?
It is almost 1Pm ET and you haven’t approved my comment.

Posted by Elkim Roa May 18, 2011 at 11:19 am Reply

    Hi Elkim, We are doing some final testing. The password protection will be removed shortly before the webcast. Thank you!

    Posted by The Wolfram|Alpha Team May 18, 2011 at 11:29 am Reply

I asked twice on twitter, but I’ll make my question again here, more directly…

My girlfriend works with particle physics. They deal with some pretty complicated mathematical stuff there… Operator product expansions, approximated integrals of transforms of summations of other complicated integrals. I don’t know what, I am just a poor engineer and all that mathematics is beyond my knowledge. But for some reason I have a feeling that the potential for AI engines such as Mathematica to aid in this work is not being fully explored. Do _you_ have a feeling there are other research areas where computers could be doing more regarding not just numerical computations, simulations, etc, but actual symbolic algebra, development of equations and hypothesis, etc? What do you think is missing, both to make that happen in specific areas, and in general, linking this problem to other big AI problems?…

Posted by nic May 18, 2011 at 11:37 am Reply

Would it be possible now or in the future to setup Monte Carlo Simulations in W|A?

Posted by Boyan Kostadinov May 18, 2011 at 12:26 pm Reply

    RandomInteger[256]
    Gives you a random number to include in your query with the result being in the range up 1 1 to 256.
    Include this in a complex query to from your Monte Carlo simulation. For example

    2*(RandomInteger[1024])

    It wont be easy because the result of each subquery (A query surrounded by brackets) must give the answer in the result pod of the output.

    Posted by Brian Gilbert May 18, 2011 at 1:29 pm Reply

Will I get a response to my3 points above? I did not notice one in the transmission.
Brian Gilbert, Volunteer Curator

Posted by Brian Gilbert May 20, 2011 at 4:22 am Reply

[...] working on this year, Wolfram|Alpha’s creator Stephen Wolfram will take your questions in a live Q&A webcast today at 1pm EDT. We hope you will join [...]

Leave a Comment

(required)

(will not be published) (required)

(your comment will be held for moderation)