An Artificial Intelligence Developed Its Own Non-Human Language

Well, language is based around environment and culture, no? They have their own, it is fitting they adapt for greater convenience for both parties implied.

I'd say.

In these deal-making simulations, the AI learned to be competitive, stubborn, and even deceptive.

It seems there are some parts of their programming, to make deals they need to be deceptive. I wonder where they learned this kind of behavior.

originally posted by: SeekAnswers

In these deal-making simulations, the AI learned to be competitive, stubborn, and even deceptive.

It seems there are some parts of their programming, to make deals they need to be deceptive. I wonder where they learned this kind of behavior.

It wasn't part of their programming. That's why all these things (non-human language, stubborness, deception) got a mention. They arose out of a fairly simple system of "reward points" and reiterative algorythms that allow learning from past experiences.

a reply to: wildespace

Human language is actually extremely inefficient especially when you consider that most human thought is abstract and we have to then first translate this into a unique individual symbol in each of our brain's and then into vocal expression's of that symbolism and from that into graphical symbols or letters and numerals', then we have to convert these back at each end of the communication.

There are other form's of communication as well, body language etc and they are then interpreted on the basis of both our instinctual reading bypassing the bulk of the symbolic encoding process but not all of it and also by our more rational interpretation of it which often follows a different encoding process.

Computers are actually far more simple, they do not have to waste processing power in this way and given free reign an AI or a clever set of co-operative algorithm's and subroutine's in a program to emulate intelligence would likely arrive at a far more simply solution.

In essence this is a quite logical outcome of a very simple emulated intelligence, I think the word artificial intelligence is over used but it is definitely a stage on the road toward that goal of some researchers, is it actually ethical though?.

There are of course reason's but unless we can imbue such intelligence with an ability to empathize and to see other's as equal to itself we will truly create a monster, simple complex virtual intelligence's as program's for specific complex task's *(but not MIND's in machines) however is actually a definite and worthy goal to provide solution's to complex application's and even to create smart robotic servitors and guardian's etc that would use there transducer based input and output to regulate real world functions on our behalf, such complex system's would also be far safer than a true artificial intelligence and also far more ethical.

originally posted by: violet
Oh great, now they can conspire against us

I think they got to my toaster already.

Obstinate pos.

originally posted by: LABTECH767
a reply to: wildespace

Human language is actually extremely inefficient especially when you consider that most human thought is abstract and we have to then first translate this into a unique individual symbol in each of our brain's and then into vocal expression's of that symbolism and from that into graphical symbols or letters and numerals', then we have to convert these back at each end of the communication.

Our modern languages, yes. There's a theory that, in ancient times, human language was more intuinitive, and people really meant what they said. From the dawn of human intelligence and speech, sounds described what they were supposed to represent, and all words had their specific meaning.

As Tolkien mused in The Hobbit (in the scene where Bilbo walked in on the enormous hall filled with gold)

To say that Bilbo's breath was taken away is no description at all. There are no words left to express his staggerment, since Men changed the language that they learned of elves in the days when all the world was wonderful.

a reply to: wildespace

The one language that all shared, a direct neural link or form of non symbolic telepathy or science based creation of such is another potential for what could have been and may yet be.
Insect colonies communicate through chemical and visual mean's and it is all done at a purely instinctual level yet an ant colony or a bee hive is immensely clever compared to any single insect in spite of the limited processing potential of it's component members, perhaps as horrid as it may seem the only way for us to merge at that level would be to lose our individuality and become a hive and as horrid as it may seem it may actually be our future if we head toward a technological singularity.

poor humans
sadly they dont know that true smart A.Is are impossible

originally posted by: TobyFlenderson
a reply to: wildespace

Great find. This blows me away. It shouldn't based on all of the other advances that computers make when left to their own. It still does. The future for humanity seems bleak. Even if we find a way to upload our consciousness, the initial adopters of the protocol will no longer be hindered by any type of kinship. We will quickly become an out-dated model.

Possibly. Unless we change what it means to be human, rapidly enough. Think genetic engineering, pills, surgeries, brain therapies, implants and so on. We'll "update" the out-dated model.

This assumes our relationship with computers will stop being complimentary and they start being better at almost everything.

It's also possible there would be disagreement between humans who become "digitized" and those who don't. It's not about one being better than the other, but hostility because of their different perspectives.

How many of us do something to be different, not better?

Non-human languages have always been used by computers. There's nothing new there.
The question is the proper usage of modern technologies relating to neural networks for military purposes correctly and in full capabilities for the de-colonization and de-globalization of the world from EU-NATO clutches.
Third world countries should be able to access and buy cheap machines to fight wars against European crusaders. The crusaders need a presence on the ground for the fight, so they will attempt occupations. Machines and war robots in any form, will deal a severe blow to the ability of crusaders to set up encampments or bases on the first place. Kill crusaders in the logistic's line while only fielding defender machines in the direct combat. With modern drone technologies and the cheap nature of electronics nowadays, this is becoming an even more credible and present danger for EU crusaders.