Slashdot Mirror

machines, no matter how complex, are a tool

there are all kinds of fun things, from a Gosper's Gun to research in neural network computing

sci-fi is great too...I just thought today about re-reading KS Robinsons "Mars Trilogy"

TFA & the "Mars Trilogy" have something in common that can help our industry save Billion$...yes that much

they both view machines from a *functional* perspective...tools that can be programmed to do tasks

In the books, AI advances realistically...it basically is a function of our computing/processing power combined with our understanding of how the human brain works...it's a logical progression

This all has to do with "teh singularity"...you're either looking for how to **evolve the human race** or solve a problem in robotics

We should fear the people who ***program*** the machines...

"It's such a simple system though! Surely it's limited to it's base rules, isn't it?"

Where would you posit that consciousness comes from?

It's my belief that thoughts and feelings (and beliefs) are nothing more than complex chemical reactions. Of course, I have no further proof of that than the fact that drinking makes you drunk and lowers inhibitions, marijuana lowers combativeness in most people, and LSD can cause psychosis.

Sure, I'm down with that. Changing the chemical composition certainly does affect the cogs in our mental machinery. Of course whether we explain these phenomena at the chemical, electrical, or structural level, the concept of emergent behavior applies. An individual chemical molecule, synaptic firing, or neuron doesn't think or have consciousness, but collectively they create incredibly complex behaviors, just as individual logic gates don't create the trajectory and explosion of an Angry Bird. Conway's game of life (a favorite programming pastime of mine in high school) and Mandlebrot sets are great, simple demonstrations of how very simple things can create complexity.

The day may soon come when a chatbot will be indistinguishable from a human operator --perhaps more interesting than most of our Facebook friends. Will it be thinking? Will it be conscious?

I wrote one thirty years ago, so yes, fooling people is very easy. Since you're a psychologist you know how strong anthropomorphism is.

(Actually, I gave it up shortly after the degree and returned to programming. Computers are easier to understand and fix!)

But you know, when David Copperfield makes that elephant disappear, it's just a trick. We're easily fooled. David Ferrucci was right -- Watson doesn't think, and a submarine doesn't swim any more than a battleship does.

But, if the criterion is moving through water efficiently, the machine wins. If an online AI communicates as convincingly as a human, does the difference matter at a functional level? The best AI efforts in the next few decades will probably be able to fool a lot of people that want to humanize the machine. But I believe that before the end of the century there will be neural nets that rival our own complexity and won't need to fool anyone. It's hard to believe, considering where our technology is currently at, but look at the amazing stuff in the sci-fi of the late 20th century that's banal reality today.

So, if a software-based brain passed a Turing duck test, on what basis would you know it wasn't conscious?

And that hits the nail squarely on the head -- that is the danger that Frank Herbert pointed out in DUNE; people using "intelligent" machines to enslave other humans. It isn't the computer that's intelligent, it's its programmer.

At the moment... But, I don't think the very first hard-AI intelligent machines will be programmed exactly. Imagine a not-so-distant future where a cross-disciplinary team creates a sub-atomic resolution perfect scan of a living human brain, then captures all the data necessary to recreate a flawless simulation of that brain. Would that electronic brain not be a conscious, thinking --yet non-organic-- thing?

Yes, you're right. TFA is rather confusing on the precise nature of the thing, but the Gemini article on LifeWiki explains what it actually is:

... Alternatively, 'knightship' may refer to any spaceship that travels in an oblique direction (not diagonally or orthogonally). The first oblique spaceship to be discovered, Gemini, was found in May, 2010 with a velocity of (5120,1024)c/33699586. In June, 2010 Dave Greene constructed the first true knightship in Life, which is based on Gemini and travels at a velocity of (4096,8192)/c35567490.

In case you missed the other person (falzer) who posted the link, the search term is "unit life cell", but the only page with an worthwhile detail I have seen is http://www.radicaleye.com/lifepage/patterns/unitcell/ucdesc.html

However falzer's link of http://www.conwaylife.com/wiki/index.php?title=P5760_unit_Life_cell because it links to other interesting things.

One of those is Deep Cell which is a version of the pattern that simulates two independant life universes with a period of only 7680.

Another is the OTCA megapixel, which is a unit life cell that is larger and slower, but makes it much easier to see the state of each cell. It can also be program to follow any life-like pattern of number of neighbors needed for birth and starvation, as well as programmed to igore any (or all) of the eight bordering cells in its determination. so you could have a pattern that ignored diagonals. Of course the OTCA megapixel itself only runs under the standard rules.

Check out the forum where it was posted: http://conwaylife.com/forums/viewtopic.php?f=2&t=399&start=0

That's the game of life forum - Conway is the guy who invented it.

It may be a popular forum, but the domain conwaylife.com is not owned by Conway.

But it seems a fairly arbitrary distinction, since that destruction is going to happen and it's not going to reverse itself.

It seems arbitrary, but given that they haven't been able to do it in the last 20 years, it's obviously not an arbitrary distinction.

Check out the forum where it was posted: http://conwaylife.com/forums/viewtopic.php?f=2&t=399&start=0

That's the game of life forum - Conway is the guy who invented it.

Reading the replies suggests this is a very, very big deal. Lots of comments like "A new age in Game of Life exploration and design!" or "undoubtedly the single most impressive construction so far in Life" seem to indicate that this is a very significant achievement.

One big distinction, is a glider doesn't take 34 million generations to replicate. In fact, the pattern itself is what moves it, so it's not really generating new gliders at all. More like a "take a pixel from here and put it here" operation, caterpillar style movement.

From the article, there are other designs that should work, but they require 10^18 generations to complete the replication, which isn't workable.

The Google Docs page with the Gemini.zip file is not allowing any more downloads right now. Here is another link with more info about Gemini and an alternate download hosted on drop.io. Follow the instructions on page 2 of the original article to set it up.

Yep, and if you read the entry on LifeWiki you would see they agree with you.

"It displaces itself by 5120 cells vertically and 1024 cells horizontally every 33,699,586 generations."

What are you talking about, non-existing? They both exist.

http://conwaylife.com/wiki/index.php?title=Universal_turing_machine
http://conwaylife.com/wiki/index.php?title=Turing_machine

http://www.conwaylife.com/wiki/index.php?title=P5760_unit_Life_cell

What are you talking about, non-existing? They both exist.

http://conwaylife.com/wiki/index.php?title=Universal_turing_machine
http://conwaylife.com/wiki/index.php?title=Turing_machine

http://www.conwaylife.com/wiki/index.php?title=P5760_unit_Life_cell

What are you talking about, non-existing? They both exist.

http://conwaylife.com/wiki/index.php?title=Universal_turing_machine
http://conwaylife.com/wiki/index.php?title=Turing_machine

http://www.conwaylife.com/wiki/index.php?title=P5760_unit_Life_cell

The Game of Life is one of the first cellular automata discovered that had simple rules but complicated behavior. The rules very roughly mimic bacterial growth. One has an infinite lattice grid, and some starting set of cells on the grid are designated as alive (every cell on the grid is either alive or dead). Each new generation is made by the following four rules: Any live cell with fewer than two live neighbors dies. Any living cell with more than three live neighbors dies. Any living cell with two or three live neighbors lives on to the next generation. Any dead cell with three live neighbors (exactly) becomes a live cell. http://en.wikipedia.org/wiki/Conway's_Game_of_Life

The Game of Life is mathematically interesting because it can be shown to be Turing complete. That is, if you have a process that tells you whether any given starting configuration will eventually dieout then you can answer whether any given computer program will eventually halt. In general, there's a theorem known as the Turing Halting Theorem which says that no general procedure exists to do that for all programs.

Prior to the work in TFA, there were known configurations called "gliders" which could replicate themselves as they moved across the grid, but they only left the same number of copies. There were also configurations which could spawn gliders (called glider guns). However, no configuration that was actually self-replicating in the sense of spawning more copies of itself was known. This work by Andrew Wade shows how to make configurations that do self-replicate. His original announcement is at http://conwaylife.com/forums/viewtopic.php?f=2&t=399&start=0 and the actual configuration can be found at https://docs.google.com/leaf?id=0B9e96aFfebqqZmY5NjBkYjctY2ViNi00NmJlLTgwZDAtNmU5OTQwYjc1OWQ0&hl=en&pli=1 Thus, this very simply system is still showing itself to have surprising and interesting behavior 30 years after the fact.

Als