Add your comments
DLS Archives
May 2012
| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | ||
| 6 | 7 | 8 | 9 | 10 | 11 | 12 |
| 13 | 14 | 15 | 16 | 17 | 18 | 19 |
| 20 | 21 | 22 | 23 | 24 | 25 | 26 |
| 27 | 28 | 29 | 30 | 31 | ||
Essential Windows Apps | Do Not Track | Microsoft Office | SayNow | LibreOffice | Zeam Android Launcher | Dead Space iPhone | Firefox 4 Mobile | Firefox 4 Release | PlayStation iPhone App | Excel Tips | Android Launcher | Google One Pass | Dead Space | Google Cloud Print | Songbird for Android | NBA Jam | Internet Explorer 9 | Windows 7 Connector for Mac | Office Mac 2011 | IE9 RC
Gadget News
- Thomson's 5.3-inch X-view 2 brings its dual-SIM capabilities, Android 4.0 to la France
- Xperia arc and neo get the Android 4.0 treatment, Sony makes good on its upgrade plans
- Xbox 360 250GB Racing Bundle totes Forza and a wheel, driving talent extra
- Amazon Instant Video streaming is now live on the Xbox 360
Reader Comments (Page 1 of 1)
(Unverified)Jun 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.