There's been an accident. The young scientist--or, perhaps, his lab assistant or friends--stands stunned. He knows he's been washed in a massive dose of radiation. He knows his life will never be the same.
In the real-world, the victims of criticality accidents spend time in the hospital. Some di... More.
"NASA to Start Radiating Monkeys," noted Chris Baker (of Wired), "The kind of headline that should be followed by 'NASA to Fire PR Firm.'"
The experiments will bombard squirrel monkeys (like the lil guy above) with radioactivity to explore the possible effects of radiation in space on human a... More.
Paul Devereux's book "The Long Trip: A Prehistory of Psychedelia" presents the fascinating story of psychedelic use before the Hoffman/Leary era that we're all familiar with. Devereux travels way back, exploring shamanism, 'shrooms in rock art, the oracle at Delphi, the pre-Incan construction of t... More.
Boing Boing guestblogger Connie Choe is a health and culture writer by day and a professional kimchimonger by night.
Whenever I'm hanging out on a chairlift I like to shout that I'm going to go die a cold, snowy death. Mostly so that if I were to actually perish on the ride down I could say, "I... More.
Kevin Poulsen at Threat Level has a great item up about the growing menace of "money mules." The term refers to bank customers who've been conned into unwittingly laundering cash that hackers have stolen from business bank accounts. The con and the funny phrase have been around for a while, but t... More.
RSS Feed
Try looking up "life, the universe, and everything." For those of you Hitchhiker fans.
Rudy, I am a huge fan (met you at a signing once in Berkeley, shared my diffusion-limited aggregation art, read your books voraciously, handed out copies of "Software" like it was a subversive pamphlet... which it is.)
So please don't take this the wrong way...
Why do technology evangelizers give Wolfram Alpha a free pass? It is quite clearly orders of magnitude less useful than Google. I have not yet seen the problem it solves. Everyone I know personally who's used it says it sucks. For even the simple computations it boasts that it can do well, it sucks awfully because its knowledge model is just not up to the job, and never will be. (I typed "murders in Oakland" and it didn't have a clue. Google, meanwhile, autocompleted my query for me and returned 2.5 million hits.)
Meanwhile, the proselytizers love it. Tim O'Reilly gushed over it. All of this seems to be about "promise". Never mind the fact that Wolfram's repertoire to date hardly inspires confidence. Mathematica itself is overpriced, and only superior to Maple if you really care about cellular automata (hey, I love them, but I've also got equations to solved). Moving Mathematica from one computer to another is a major headache: you have to go through corporate channels to get a re-licensing. It BITES ASS.
Wolfram's other main opus is "A New Kind of Science", which is hideously self-aggrandizing to most people who actually *do* science, and is mostly a collection of coffee-table images plus a proof that a particular 1D CA is Turing-complete.
Why is no-one pointing out that the Emperor is stark bollock naked? Google has been around for a few years now and has a lot of challengers. Can we just agree that Wolfram Alpha is CONSIDERABLY less use than Google, and for the foreseeable, always will be? And in fact, that it's currently WORSE THAN USELESS? (One counterexample of a successful use case please.) Or do we have to suffer more of this tedious everything-new-is-wonderful "it is expected to be really useful when it improves at some indefinite point in the future" hype?
Completely failed on 'Ukrainian rent boy'.
Ian---I'm posting Alpha as a matter of interest, not necessarily because I think it has reached its potential. It's a matter of public interest.
As for "A New Kind of Science"---I'm weary of arguing with people about it, but I happen to think it's a great and important book. People who disparage it tend, I suspect, not to have put in the effort to really understand the ideas being put forth.
I went into some detail about the book in my review in the American Mathematical Monthly, November 2002.
The book did revoutionize my way of seeing the world, and I am (some of the time) a scientist. As you probably know, I even wrote my own version of "A New Kind of Science," my tome, "The Lifebox, the Seashell, and the Soul."
And I think there are quite a few scientists who would disagree with you about Mathematica.
So---in short---yes, Wolfram's style can be off-putting, but his ideas are worth some attention.
Wolfram claim that they own the results of the computations their engine performs, and require attribution. That alone is enough to make me not use it. http://tech.slashdot.org/article.pl?sid=09/05/19/1846258&from=rss
Wolfram Alpha is NOT Google.
Let me say that again.
Wolfram Alpha is NOT Google.
If you try to use it like you use Google, well, you fail.
For the majority of the unwashed internet masses, there's NO reason to use this site.
Here's a little experiment for you Ian: Enter "sin(3x)" into Wolfram Alpha and Google. Compare the different result sets.
Antinous, be real. It is designed for quantitative queries, and does not know many synonyms.
Try "number of rent boys (whose nation of origin is Ukraine (the former Soviet Republic)) whose services can be procured for X dollars, where X is the average monthly stipend for a columnist at National Review, the Weekly Standard or the Washington Times"
It always sounds to me like the search engine they'd use in hell (cf Wolfram and Hart)
Meh.
Pure hype.
It is not even able to display SDS for chemical compounds, and it is absolutely useless in Biology.
It has a couple of nice features, like charts and stuff, but the inability to get the data or go to the source of the information completely sucks.
It is like talking to a very, very, very ignorant idiot savant.
Rudy: Agree with Ian.
Also, I spent many, many hours with a new kind of science. if it doesn't apply to your kind of science, it just isn't that useful. I think the main issue I had with it is the broad sweeping generalization it made about the extension to every kind of science. I personally did not find it to be true.
Also, 'number of rent boys' caused it to freeze like a madman. Not sure what a rent boy is but perhaps it is contacting the FBI? Thanks a lot, guys.
So, Wolfram is offering for free what you used to have to buy Mathematica to do? Makes perfect sense.
@Ian: Your comparison with Google is a good one, but in my mind illustrates the importants of W|A. Do you really want to search through 2.8 million hits on "murders in oakland"? Sure, W|A doesn't have that dataset (yet), but it did identify that the Oakland in California was the most likely city you wanted to filter on.
And once they do add this data to W|A, you'll be able to ask for "What is the murder/suicide ratio in Oakland, plotted over time compared to cloud cover?", something that Google will never be able to give you an adequate answer for.
The root difference is this: Google does keyword searching, W|A allows you to identify the links of data and do inteligent pivoting on that data.
@GuidoDavid: I'm looking forward to their exposure of the API. Imagine an Excel spreadsheet that lets you access the world's knowledge, and pivot anything by any other thing, or gives you access to the raw underlying data to display in your own presentation format. It's wonderous for discovering patterns. I'm personally more excited about W|A than I was about Cyc0, as W|A seems much more practical.
@Rudy: I'm surprised you didn't link the screenshot of the BoingBoing HTML parse, or a graph of the estimated BoingBoing traffic for last year. :)
In the end, this is the first release, and the data they've entered is remarkable to me. I'm a bit taken aback by all the haters. Where's the unabashed optimism that the future dreamed about is finally starting to peek around the corner?
Maybe I'm the only one that had those dreams.
Or the only one that's kept them alive...
The best use it seems to have is as a really advanced calculator sinc x * sinc y #11030595112785124913 or say seeing the Heart Surface
Rudy @4: I'm not saying Wolfram should be ignored! You're quite right that many scientists would prefer Mathematica over Maple, and it was revolutionary in its day.
My view (not specifically directed at your post, but this seems like a fair enough place to air it) is that I do not think any new search engine, post-Google, warrants the kind of uncritical excitement that seems to have accompanied the release of Wolfram Alpha.
I too tire of the arguments over A New Kind of Science. Many initial responses were very skeptical and I agree it deserves a fairer look. Fundamentally I think it is a pop-science book (which is fine -- popularization of science is difficult, important and wonderful) as opposed to a fundamental technical text (notwithstanding occasional rigorous arabesques, such as the Turing completeness proof). A popsci book can blow your mind, but it should not claim equality with Principia Mathematica.
Yes, this is a stylistic criticism and should not detract from the strong ideas in ANKoS. But nor does a florid style remove the need for a serious critical examination of all the ideas. Thus my irritation at the blindly optimistic & premature Wolfram Alpha hype (present company excepted because, Rudy, the day I lapse into irritation at you is the day my brain deserves to be eaten by crazed ice-cream vendors while my body is devoured by self-assembling robot ants)
Seems to be quite ggod at science that I don't do, but for Cryptography and discrete mathematics it is useless - it can't factor the product of two large primes or do the discrete log problem or even find pre-images for hash codes ;).
I like the result for 88mph though http://www57.wolframalpha.com/input/?i=88mph
Just wait till it gets to Beta :)
MrScience @12: no offense, but come back to me when WA actually can compute "the murder/suicide ratio in Oakland, plotted over time compared to cloud cover". Until then it is vaporware.
There is a basic problem of knowledge representation at the core of this. Curated, comprehensive knowledge bases are a fundamentally pre-web way of doing things. They are always miniscule and pathetic compared to the anarchic free-for-all of freeform text corpora.
Compare Encyclopaedia Brittanica to Wikipedia, as e.g. Nature did.
Yes, unstructured datasets are also harder to compute on, but this is not news. There is a long history of research (and failure) in this area and it is deeply irritating to see so few questions being asked of Wolfram Alpha as they arrogantly (and speciously) announce that they have conquered these wilds.
The hard questions that reporters should be asking: how does Wolfram plan for maintenance, turnover & expansion of their curated knowledge base? How will they pay for the army of curators required for this? How will they prioritize new entries? Who is actually using this on a regular basis?
Proportion of sentences in "A New Kind of Science" that start with "So", "And", or "And so":
Answer returned: "Stop picking on my writing style!"
Seriously. The whole book is "And we can see that numbers are not what we thought they were. So it is apparent that we need to rethink what they are. But this will not be easy. And so that's why you need to read this book." Gah.
A thought:
Using Google and Wolfram Alpha together is useful. They both make each other more useful.
There's no use in comparing the two of them; they are different things with different purposes.
Anon@6:
Here's a little experiment for you Ian: Enter "sin(3x)" into Wolfram Alpha and Google. Compare the different result sets.
so you're saying it should not be compared with a search engine (e.g. Google, 1995) but, rather, with a graphing calculator (e.g. Casio fx-7000G, 1985).
I think I'm still missing why my mind should be blown.
PS Anon@8: Wolfram & Hart, LOL
I found Rudy's review of A New Kind of Science. http://sjsu.rudyrucker.com/~rudy.rucker/wolfram_review_AMM_11_2003.pdf
I must say that I don't really know enough to know how much is new, or how much of Wolfram's speculating may or may not in fact be true. But I did find the book very interesting both times I read it, and I think it's nice that someone else felt the same way. I also read Rudy's The Lifebox, the Seashell, and the Soul, which talked about some of Wolfram's stuff and was very interesting as well.
BTW, I'm not saying that curated databases never work, and "pre-web" may be a stupid generalization. A great example (from bioinformatics) of a curated controlled knowledge framework is the Gene Ontology. But most people now acknowledge that curated databases have fundamental limitations compared to freeform text. And I don't think WA is anywhere near as mature (not to mention domain-focussed) as the Gene Ontology.
Brandon @5: Wolfram claim that they own the results of the computations their engine performs
They sure as hell don't own my Ukrainian rent boy.
We are not haters, come on!
I was only expecting something that actually stood to the expectations, not something that does not even has access to GenBank!
The concept is neat, but again, being unable to collect the data is plain silly.
Wolfram Alpha is going to completely change the way we build our cities.
I asked it if God existed and it showed me a picture of a Border collie to prove it.
At least both books changed the way I think about things.
"Interpreting Boing as Boina" is a Mathematica computation?!
OK, so I typed "Integral of sin(x) from 0 to Pi and it gave the answer as 2 and drew a pretty graph.
I tried "degrees of freedom" expecting a mathematical or physical interpretation and it told me how many bachelor's and master's degrees were awarded at Fairleigh Dickinson University. A work in progress.
If anyones interested, it's available as an addon on Firefox already.
WA does have some nice features for non-scientists. I like the natural language search / calculation combination.
For example, if you search for "weight of 3 cups of flour" it'll tell you the volume of a cup (with options for US, Imperial or Metric cups), the density of flour and then calculate the weight.
A couple of clicks further gave me the full nutritional information for three cups of flour; typing in a full recipe gives the aggregate nutritional data for that meal.
Out of curiosity, it also correctly answered "(distance to pluto) / (speed of light)" to give me the current distance to pluto, the speed of light and thus the time light takes to get there, expressed in a variety of units. An astronomer friend of mine confirmed that the WA's value for "current distance to pluto" was exactly correct for the day I asked. Going back to the page, I found commands to calculate all manner of interesting astronomical data.
The point of WA is that, while google can generally only point you to answers that a human has already written somewhere, WA can calculate new answers to questions about anything in its database.
I see this as a bit of a catch-22. It will only become a powerful tool when organisations (govt statisticians, public science) start releasing data sets in a format that WA can gobble up. But those organisations will only start spending resources to do this when WA is already a powerful tool.
The complaints about its natural language processing are valid, but it's a bit like google: its users need to learn what kind of questions it's good at, and how to construct their search terms.
Of course, it has its foibles. May favourite examples are the searches for Wales and Scotland: Instead of small countries with a few million people each, WA thinks they're English towns with a few thousand inhabitants.
The biology tools are weird too. It'll translate a DNA sequence to protein and see if it matches known genes. If you give it a gene name (try IFNB1) it'll give you some nice information too. This is fine, although nothing that isn't already easily available. The weird thing is that this only seems to work for human genes, while the databases I assume it's pulling this from contain much, much more.
It sure nails this question.
See?! It's USEFUL!
I've a question about source
I tried to find some useful data about my country and Wolfram gave me an interesting statistic which I can't find anywhere else. Under "sources", the primary data source is "Wolfram curated data". What exactly does this mean?
Under background data, there're names like WHO, NCHS, etc. Does Wolfram find some kind of average?
I asked it to give me a measure of Wolfram's arrogance in units of Golden Gate Bridges, and it said 1729.
As others have pointed out, if nothing else use it as a really great web-based calculator. For example: 5/1542
Sure, that's fine, but also try:
55*(plank constant)/(sin(.23))
Or:
integrate (x^2)*sin(x)
If it isn't about numbers then wolfram alpha doesn't do it. Even if it is about number it might not do it. The biggest weakness is that there's no way that I can see to correct it. It doesn't think the small town I'm from exists and there's no way I can see to refine the search to see if maybe it does indeed have some data about my town.
It's surprising, but I find that I receive a little comfort knowing that I have immediate access to the repeating continued fraction expansion of any square root.
e.g. sqrt(161) into Alpha yields: [12; 1, 2, 4, 1, 2, 1, 4, 2, 1, 24]; and indeed, root 161 equals 12 plus one over (1 plus one over (2 plus one over (4 plus one over...))).
There's a "Give us your feedback" link at the bottom of every WA page; I've already fired a few things in there.
If it knows 'life, the universe, and everything' is 42, then it should know the temperature at which books burn, for example... (451 F, per Ray Bradbury).
Interestingly, plugging "42" gets you no Hitchhiker's references, but 'life, the universe, and everything' gets you '42'.
Not being a scientist, mathematician or statistician, I'm not sure how much use WA is for me. Interesting, though.
It's got some issues.
A search for "queens new york" yields...
"Assuming "queens" is a government position | Use as a university instead
Input interpretation:
queen of New York (US state)
Interesting indeed, but as everyone has pointed out already, this is practically useless. Unless the word or phrase sought is vaguley similar to SOMETHING in or about the United States it just gets confused.
I want to be able to let it know it's wrong and that it should look other other possible meannings.
"The biology tools are weird too. It'll translate a DNA sequence to protein and see if it matches known genes. If you give it a gene name (try IFNB1) it'll give you some nice information too. This is fine, although nothing that isn't already easily available. The weird thing is that this only seems to work for human genes, while the databases I assume it's pulling this from contain much, much more."
Hmmmmmmmm.
Never tried that.
When I typed "Hexokinase" and I got no result at all, I assumed it was worthless. Anyway, there are a lot of enzymology databases it could use for that.
For example, if you search for "weight of 3 cups of flour" it'll tell you the volume of a cup (with options for US, Imperial or Metric cups), the density of flour and then calculate the weight.
However, if your query is "highest mountain in West Virginia" (a question entered in beta), it still won't know what to do.
Try this. You'll be disappointed:
Notable event for July 20, 1969:
I registered for API access, but it's apparently not open yet -- in fact, looking at their API document it seems to be half-baked at best, just like the search results.
One thing I did like about the API is the new "validatequery" function:
So Wolfram clearly knows Alpha cannot return sensible results for many (most) queries. Nice of them to make this available as a boolean API call!
After a few minutes browsing, it looks like W|A's competition isn't google, but wikipedia.
My favorite experiments give up on artificial intelligence and work at making collaboration between human intelligences easier. The idea of holding some massive database without relying on input from the audience seems like a throwback.
To perfect W|A, I'd let users tag and edit content. But, then, Ward Cunningham already did that.
It still has no idea what "British Columbia" is.
What's up with that?
Yep, wikipedia:
knows what's the tallest mountain in WV
Has a full description of hexokinase
Gives a complete description of the sinc function
Has conversion factors between a cup and anything else
In short, everything W/A has done successfully so far, wikipedia does better. W/A looks to me like a web calculator: if they had focused on that, they'd have a real interesting product.
I think the real point of comparison is that, so far, the coolest thing W/A does (that I've seen) is know that 88MPH means Marty McFly, and 42 is the answer to everything. Honestly, these things would probably be considered graffiti in wikipedia, and would be deleted (or properly linked) immediately.
@Bugs
I similarly used it to work out how much sugar is in a tablespoon. Granted, google returned similar results but the top answers were from Yahoo Answers.
Also, typing "hello" gives you a nice response, whereas typing "hello" or "hello google" gives you nothing relevant. I am somewhat disheartened, although my first thought was to google-bomb the phrase so that the first hit was something along the lines of "fuck off" (with the GIS result being the finger of course).
I think it is an interesting coincidence that this came out relatively soon after microsoft announced they were killing encarta. Just as Encarta was an attempt at a digital encyclopedia, this is an attempt at a digital almanac.
Count me among the W|A skeptics. Remember, the idea is to "make all systematic knowledge immediately computable and accessible to everyone." As far as I can tell, by "all systematic knowledge," all he really means by this point is "mathematical calculations." I've personally tried several different questions involving computations of real-world data, and have come up with absolutely nothing useful.
Good idea. Maybe someday it'll be useful.
A bit of work on W/A makes you realize what a great software-as-service version of Mathematica it could make.
The world really needs some hardcore math software-as-service. Something that made it easier to create, share, and manipulate math formulas and graphs. Like a collaborative LATEX editor plus visualization and data sharing. That'd be great!
Maybe Google docs will get there some day...
Anonymous @40, is this query non-United States enough for ya?
I wolframalpha'd "boing" and it said highest elevation 300m. So it knows that most of the world uses metric.
I'm looking for something like Wolfram Alpha where I can ask it any question and it will return a correct, verifiable answer. I would also like it to see into the future and the distant past, and preferably deliver its answer in a thunderous and booming voice, preferably the user-interface would be a giant stone face with a gaping mouth and scary eyes.
http://yoshout.files.wordpress.com/2008/08/zardoz3.jpg
yeah.
Anyways, need the new word; wolfa'd? frammed it?... wolfram is tungsten... going to have to come up with it soon... suggestions?
I really like the google/wolfram mashup addon for Firefox.
Agh please!
Stop complaining - WA pushes the frontiers, cmon. Google is good, but works on basic semantic principles. WA works in a more focused way to deliver what you ask for - try "weather in New York" - all sorts of useful, straight away. Google needs trawling, popping in and out, annoyance and time.
We're slowly getting out of this basic search engine world (of which Google is clearly the best), and into smarter software that answers our questions more readily.
It's like so many online / software apps out there now - they work very well at this "simple but gargantuan" level, but don't tie things together in the way that our minds work.
I applaud WA and all it stands for, hope it's a goer.
It's broken.
My GDP related queries all failed.
Yes, but is it...alive?
I've been living for that phrase lately, sorry.
Ok, I am totally looking forward to further refinements of this gizmo. No, it's not perfect; I asked it what I should eat for breakfast and it failed me utterly. That aside, everything good has to start somewhere relatively simple.
Wolfram's ANKoS is absolutely a worthwhile book. It had an undeniable impact on my work as a painter. Deprecating the work of a scientist/philosopher because of his less-than-lavish prose is really about the most laughably absurd critique imaginable.
"waa'ed?",,,no, I think not.
even with it's limitation that it only knows about human genes - I thought, ok, let's let it do some math over that data.
it couldn't deal with "average length in bp of human genes"
#62, try searching only "dna".
If you're into genetics or another field of expertise you could check out Participate in the Wolfram|Alpha Project.
#54 Nesbitt --- Awesome. I would use your system even if it had the same limitations as Wolfram|Alpha.
#57 - I don't live in New York, so I tried "Weather in Woodland Hills," which is where I do live, instead. Unfortunately it returned me info on Woodland Hills, Utah, whereas I live in Woodland Hills, California. Google, on the other hand, instantly gave me the info I wanted for California with the same query. Winner: Google.
A check on Wikipedia shows that Woodland Hills, California, has about 70 times the population as the town with the same name in Utah. Why W|A's first assumption was Utah, only Wolfram knows.
@64 W|A uses geolocation based on IP... maybe the public IP you were using came out closer to Utah?
Typing in Tides yesterday correctly identifies that I'm in the Seattle region, and provides the correct tides for this area. Asking it for the tides on December 1st, 1600 results in it doing the various orbital calculations for that date, and provides them for my location.
Let me see Google/Wikipedia do that... Oh? You say that this isn't a real question? Hmmm.
Sorry if this comes across as antagonistic. I'm just completely blown away by the cynicysm here.
wolfa.com is an illegal mirror site. It is not wolframalpha.com. If I were you I would be particularly wary of their downloads page.