Unix and Linux network configuration. Multiple network interfaces. Bridged NICs. High-availability network configurations.


Reviews of latest Unix and Linux software. Helpful tips for application support admins. Automating application support.


Disk partitioning, filesystems, directories, and files. Volume management, logical volumes, HA filesystems. Backups and disaster recovery.


Distributed server monitoring. Server performance and capacity planning. Monitoring applications, network status and user activity.

Commands & Shells

Cool Unix shell commands and options. Command-line tools and application. Things every Unix sysadmin needs to know.

Home » Featured

AI Research: Playing Mind Games

Submitted by on April 23, 2008 – 3:26 am 2 Comments

The news of Artificial Intelligence research always attracted my attention. I myself was a fan of new technology, especially I am a gamer as well and my friends, looting and trading diablo 2 items for ages now, I would like to see things like this because of its potential to be one of the best gifts for gamers like me, having something that equipped with Artificial Intelligence is cool! Whenever this subject comes up on Slashdot, I just can’t resist posting a comment. The desire usually is to dump a bucket of cold facts onto the fire of wild imagination. The recent such occasion presented itself with the discussion of the “Two AI Pioneers. Two Bizarre Suicides. What Really Happened?” – an excellent story in the Wired magazine on the recent suicides of two well-known AI researchers.

Chris McKinstry and Pushpinder Singh – a self-taught AI enthusiast and an MIT researcher – both had grandiose plans of building giant databases of trivial knowledge that would be used in creation of artificial intelligence. This artificial intelligence refers to AI that can be used in casino sites and bingo sites such as UmBingo, to ensure a fair and transparent online casino and bingo gaming.  Both men ended their lives with a gas-filled bag over their heads. They were in their thirties.

Computer technology made tremendous advances in the past twenty years. It includes the creation of revolutionary online casinos that brings real casino online, 666 Casino is an example of this, this kind of website allows people to play casino online and win real money online. These remarkable achievements make us believe that nothing is impossible in the world of computers. What remains for us to get better, is be patient and read reviews of slots games such as the white wizard to fully understand how the game works and then develop a strategy. No self-respecting sci-fi author will publish a story if it doesn’t have a thinking, learning, talking computer. HAL 9000 of Arthur C. Clarke’s 2001: A Space Odyssey or the Enterprise computer from Star Trek: The Next Generation – we are all well familiar with the concept and even got used to it. Not just sci-fi fans but many serious researchers believe that true, human-like AI is achievable in the foreseeable future.

Singh and McKinstry both believed that the key to creating AI is a vast database of trivial facts: dogs have four legs, sun is hot, ice is cold, etc. Hundreds of millions of such statements would be available to the computer. This was supposed to be the foundation of machine intelligence. A preposterous idea, if you ask me: knowledge is neither a prerequisite for intelligence nor the source of it. A newborn child possesses none of these facts and yet he is already intelligent.

Intelligence is an ability we are born with. It’s an ability to turn information into knowledge. If someone is an idiot, forcing him read an encyclopedia is not going to make him any more intelligent. The only form of intelligence we are familiar with (and, therefor, the only form of intelligence we can hope to artificially recreate) is our own – human – intelligence. But how can we even talk about building AI if we have no complete understanding of how the original works?

Imagine if you showed a computer to a craftsman from the Victorian era and asked him to make a copy of it. He would have carefully examined the machine and built an exact replica, accurate down to the minute visible detail. It would have looked very convincing – a keyboard with buttons you can push, glowing “LEDs” with candles hidden behind tiny glass windows, and IDE cables made out of sheep skin stitched together to resemble the originals – but, obviously, it would have never worked.

This is what we are trying to do today in the AI field. We don’t understand what we are trying to copy; we don’t know how it works; we are just beginning to learn how it’s put together on a physical level. Some propose building an extensive neural network; others say that building a massive computer resembling the internal structure of the human brain is the answer; and researchers like Singh and McKinstry thought a huge database of trivia was the key to unlocking the secrets of AI.

These different approaches to AI have one thing in common: they rely on a certain element of spontaneity. It is implied that something will happen at some level of these massive and complex systems that will give rise to intelligence. And, of course, there are charlatans who write simple software that performs simple everyday functions – like the Rhoomba vacuum cleaner stumbling around the room and learning its layout, while cleaning your floor – and claim that their gadgets are “intelligent”, simply because they don’t run into the same coffee table twice.

Some of these “intelligent” systems are complex and convoluted enough to briefly fool you into believing that you are dealing with artificial intelligence. However, any attempt at in-depth interaction quickly clears the fog of confusion. A state-of-the-art supercomputer can beat a world chess champion, but it can’t maintain a convincing conversation with a three-year-old child. Why? We know and understand the rules of chess. We build a supercomputer for the specific and narrow task of following these rules. There are no rules for talking to a three-year-old: he can ask you a deeply philosophical question or tell you that he pooped himself.

Skeptics of AI ask: how can one program intelligence? Modern software claiming some AI characteristics is written by programmers using many of the standard programming tools used for more down-to-earth tasks, like writing word processors and computer games. Anyone who ever played an on-line computer game with the help of P4R Gaming boosts against real-life opponents will tell you that no “AI” comes close to replicating the experience, even the League of Legends 2017 worlds skins look out is way better. For more details, this page will tell you all you need to know.

A program is a series of steps – the outcome depends on the input, but can always be predicted regardless of the complexity of the code. Programmers sometimes talk about unpredictable results they get from their software. However, these results are unpredictable not because they could not have been predicted, but because the programmers failed to predict them: they failed to test the software under some specific combination of inputs. These unexpected behaviors are what we call “bugs”.

On the other hand, the blueprint for our own intelligence is programmed within the DNA. Some AI researchers argue that, if it is possible to code human intelligence onto the DNA, then it should be possible to program artificial intelligence. This is a strong argument, but it sidesteps one important point: the DNA is not intelligent on its own. It’s a carrier of design specifications – not the actual “hardware” that runs the program we call “human soul”.

We don’t know what makes us intelligent. Just a fraction of our DNA makeup separates us from apes and over 40% of our DNA is identical to that of a fruit fly. Apparently, the devil is in the details. If we could build an exact electronic or software copy of our brain – as far as we can see it – and mimic its physical functionality down to the last neuron, will it become spontaneously intelligent and call you “mama”? Probably not. This would have been too simple. Creating AI in the true sense of this term – as artificial intelligence equal to our own – would be a monumental achievement comparable perhaps only to making a contact with an extraterrestrial civilization.

Do I want a laptop that’s smarter than I am? Probably not. Besides, it would be useless: why would it ever want to browse porn sites? But somehow I don’t think that this is something I would have to worry about in the foreseeable future. AI is not like creating a radio, or an airplane, or an A-bomb, or going to space – it’s way out there, beyond the Moon and the Mars and beyond our lifetime.

Print Friendly, PDF & Email


  • sarah w says:

    The area is not matted so please sensible ones thanks.

  • Travoiz says:

    So I’ve been wanting to do this for a long time now. Build an AI. Not like something like a video game but an AI which I can use to do tasks on my computer for me. I’m currently on a mac, just to let you know. I also want it to be able to speak, learn, and do what I ask it.

    I have a 3d model I would like to have the AI use, but I don’t really know how/what I would use to build it. I know a small amount of HMTL code, and I have a couple books on coding. Can anyone help me?

    Thanks a ton!
    If you would like to email me on this topic, please:
    I know building AI is very, very complex. I am willing to spent as long as it takes to build it.

Leave a Reply

%d bloggers like this: