Slashdot Mirror

Good post, man, but

We do not expect a computer to read our ordinary handwriting off a piece of paper.

Lola Cañamero's Emotion Page

Steve Allen's Home Page

Crikey, you figured that out after two semesters. I guess I wasted 4 years of my life doing a degree in it all then... I must never have cottoned on to how well expert systems such as Mycin and Dendral actually perform.

You think programming is just the "intelligence of the programmer"? Guess again -- many people have AI systems running which program themselves, coming out with emergent behaviour which the programmer never expected.

Do you really think that a person can simplify circuit boards to their simplest form by themselves? I thought not. I know that Julian Miller can't, but that using his Cartesian Genetic Programming he's managed to wirte programs that do just that. Thus proved that a computer program can ultimetaly be more than the sum of its external inputs.

Crikey, you figured that out after two semesters. I guess I wasted 4 years of my life doing a degree in it all then... I must never have cottoned on to how well expert systems such as Mycin and Dendral actually perform.

You think programming is just the "intelligence of the programmer"? Guess again -- many people have AI systems running which program themselves, coming out with emergent behaviour which the programmer never expected.

Do you really think that a person can simplify circuit boards to their simplest form by themselves? I thought not. I know that Julian Miller can't, but that using his Cartesian Genetic Programming he's managed to wirte programs that do just that. Thus proved that a computer program can ultimetaly be more than the sum of its external inputs.

the mindsweeper game never has any thing that you HAVE to guess at

Given that Minesweeper is NP-complete, are you so sure of that assertion? In a sufficiently-crowded field, you almost always get to some point where you can't deduce from the surrounding squares whether or not there's a mine in a space. You end up guessing and hoping for the best when this happens. I suspect that the Minesweeper where you never have to guess isn't the true Minesweeper.

That is untrue. Solution space exploration would be critically curtailed with no mutation. In fact, plain vanilla evolutionary strategies employ only mutation for solving problems.

Also, mutation can be a more potent exploratory force than cross-over.

The moral of this story is that you should prefer canonical sources for knowledge than quips made on /.

I looked up Klatt, like the AC mentioned - here are some links for the rest of us...

GPL'd Klatt Synth Source

RSynth Speech Synthesizer - Klatt based synth - go to /soundapps to download gzipped code

KPE80 - A Klatt Synthesiser and Parameter Editor

Worldcom - Generation Duh!

And for once it's not because I was sleeping in lectures ; )

I'd suggest that NASA isn't going to put the rock samples that it may one day bring back from Mars at a great expense into an incinerator.

Also, ...it has been found to remain infective after 360 degrees C for 1 hour or even after incineration....

Samples will need to be studied with Level III[search prion] precautions. I trust if there is a possiblilty of TSE contamination, NASA will quarantine. I believe there is a possiblity that TSEs exist on Mars, and I support the panel.

Looks pretty cool stuff... I wonder if there are many places doing stuff like this.

Looks pretty cool stuff... I wonder if there are many places doing stuff like this.

Looks pretty cool stuff... I wonder if there are many places doing stuff like this.

The allure of robot wars/battle bots/whatever is not watching them maneuver or negotiate obstacles; it's seeing machines get mashed! Just like with Motor Racing, there is an unwritten rule; audiences will pay to see destruction!

The truth of the matter is that most competitor-built robots are pretty useless and inflict little damage on others robots, therefore, to up the 'destruction quota' Robot Wars has 'House Robots' that will beat up on any robot that breaks or ventures into the wrong area of the arena!

Also, I read a post by someone who thinks watching autonomous robots fight would be fun.
[BZZZT] Wrong.
Having actually built several autonomous robots I can tell you that it would be the most BORING program ever to grace the beloved CRT. why?
1. The sheer difficulty in building a machine that can drive itself is staggering! Car engineers have tried this for decades without success.
2. Once you've got it to maneuver, how does the damn thing know where the opposition are? let alone where the walls and obstacles are! How do you tell the difference between robots and obstacles? Mind-bogglingly difficult.
3. Assuming you got your robot to do the above, you then actually have to have a reasonably destructive weapon in order to win! and use it correctly! How do you test this? You cant just turn on this autonomous, chainsaw weilding maniac in your living room!!!

At the end of all this, you'd have a program that involved lots of huge, hideously complex machines mercilessly attacking the walls, floors, etc. A highlight would be when one actually drove in a straight line before deciding that an object in the distance was actually an enemy and proceeding to kill thin air. Admittedly it might be interesting for Software people to try and analyze the logic beneath their behavior, but for Joe Shmoe it'd be dull, dull, dull.

Asimov would not be impressed.

Jan.

Some could say Comp. Eng. is more market driven and Comp. Sci. is more research driven, but at undergraduate level they're really the same. It's almost impossible for an undergrad to participate in research -- they just get taught how to build software and understand methodologies.

I take an Atrificial Intelligence degree, which is definitely science... but software engineering and derivatives could definitel be classed as engineering.

There was a story on Slashdot recently about a Turing Machine implementation in Minesweeper. The paper is here (in yucky pdf format).

His proof is very similar to the Life proof -- which makes sense, because when you think about it, Minesweeper is a lot like Life. (That's 'Life' with a capital 'L'... I'm not trying be profound here.)

to the Millennium Prize Problems page
to Ian Stewart's article on the problem.
to Stephen Cook's mathematical description of the problem (in .pdf format)
to Richard Kaye's Minesweeper Page

More details of the maths involved can be found at The ClayMath Institute's webpage and some related papers at R.W.Kaye's webpage

The author of this paper has a web page here:

He also has a page specifically about this Minesweeper business here.

I like the other paper proving that minesweeper, with a little variation, on an infinite board, is Turing-complete. Fun! :)

---

The author of this paper has a web page here:

He also has a page specifically about this Minesweeper business here.

I like the other paper proving that minesweeper, with a little variation, on an infinite board, is Turing-complete. Fun! :)

---

I did find an interesting resource at http://web.bham.ac.uk/M.L.Wyszynski/people/publwys z.html for the feasability of such systems as hydrogen gas enriched fuel-based systems.

I'd personally love a car that ran on something other than gasoline, if only for the pure novelty of it.

Compression CANNOT guarantee anything better than 1:1 ratio - it is ENTIRELY dependent on the data.

- The US patent office no longer grants patents on perpetual motion machines,
but has recently granted a patent on a mathematically impossible process
(compression of truly random data): 5,533,051 "Method for Data Compression".
See item 9.5 of this FAQ for details.

As can be seen from the above list, some of the most popular compression
programs (compress, pkzip, zoo, lha, arj) are now covered by patents.
(This says nothing about the validity of these patents.)

Here are some references on data compression patents. Some of them are
taken from the list ftp://prep.ai.mit.edu/pub/lpf/patent-list.

....
9.2 The counting argument

[This section should probably be called "The counting theorem" because some
people think that "argument" implies that it is only an hypothesis, not a
proven mathematical fact. The "counting argument" is actually the proof of the
theorem.]

The WEB compressor (see details in section 9.3 below) was claimed to compress
without loss *all* files of greater than 64KB in size to about 1/16th their
original length. A very simple counting argument shows that this is impossible,
regardless of the compression method. It is even impossible to guarantee
lossless compression of all files by at least one bit. (Many other proofs have
been posted on comp.compression, please do not post yet another one.)

Theorem:
No program can compress without loss *all* files of size >= N bits, for
any given integer N >= 0.

Proof:
Assume that the program can compress without loss all files of size >= N
bits. Compress with this program all the 2^N files which have exactly N
bits. All compressed files have at most N-1 bits, so there are at most
(2^N)-1 different compressed files [2^(N-1) files of size N-1, 2^(N-2) of
size N-2, and so on, down to 1 file of size 0]. So at least two different
input files must compress to the same output file. Hence the compression
program cannot be lossless.

For data compression in memory to succeed, you MUST have an option to cache the "extra" memory to a swapfile incase the prediction logic fails and you run out of physical ram. If you do not, you will tank your system, bigtime.

Sorry, but I'm very leery of any "memory compression" - it requires OS support to function. Period. You aren't going to just plug in a miracle DIMM and make it work. I hope IBM is opening the spec (it looks like they are) and that OS development people quickly embrace this, or their hardware will take a nosedive in the market.

I don't know if this is the same thing but a couple of universities in the UK are working on 'gravity wave' telescopes at the moment. This page contains some information on the group at Birmingham University and gravity waves but does not mention the telescope. However when I was there last autum they showed me the preliminary ideas for one.

Would the next generation programmers write in "logic language" instead of C++?

Most likely not - but automatic verification of programs using logical constructs is a big growth area.

You can test a program with all possible inputs, and have a clean run. But this does not mean the program is 100% reliable. You must prove the program is correct if you want to be sure it is good enough for circumstances such as shuttle or aeroplane flight.

With all the complexities of semaphore control in parallel computing, you really have to make sure a program enters and leaves critical sessions at the correct times, without anything else running (that has been designated mutually exclusive).

Many expertes believe that some Airbus crashes were caused by incorrect verification.

On a single processor machine, this is much easier, but how many space shuttles do you know of that only have one CPU!

Have a look at some of the links at Dr. Mark Ryan's page (university of Birmingham) for some more info.

Well, Poplog includes tools for Common Lisp, SML, Prolog, and POP-11, all of which can be combined and work together. It's available for both Unix (including Unix) and Windows, although the Windows version is not as good.

This probably isn't what you're looking for, but it allows you to combine Lisp and ML.

~~~~~~~~~
auntfloyd

AC: Most deaths are caused by sky-diving accidents.

PB: No, most deaths are caused by heart failure.

AC: Um, and just what do you think it is that sky-diving accidents do?

Let me repeat: most cancer isn't caused by viruses (although some is). The top two causes are smoking and diet, although people will claim that just about anything is a cause, including stress, premarital sex, and snakes.