What would AIs call themselves?

[Celebrim]

Celebrim's Laws of Ethical Machine Sentience

1) Thou shalt not make an unhappy machine.
2) Thou shalt not give a machine more sentience than is required for its duties.
3) Thou shalt not make a machine which can be unfriendly to its owner or culture.

Sentient AIs are formally refered to by abbreviations which accord to the general paradigm of behavior that they were intended to have, for example 'vecs' for robots that correspond to Hans Moravecs vision of intelligent computing or 'movs' for robots of an Asimovian character, 'rosses' for those of a more Strossian character, and so forth. In general, a robot of unknown classification is generally referred to as a 'bot'. All apparently or provably sentient bots are entitled to the pronouns 've' and 'vem' as opposed to 'it'.

 

[interwyrm]

I have to disagree with Celebrim's diatribe against humanlike robots... but I do like his naming conventions. Giving a nod to science fiction writers for considering the ethical implications of technology before the technology is actually available is a fantastic idea.

I'd like to propose my own solution as well. Just as Aldous Huxley uses different designations for different breeds of humans in Brave New World (Alpha, Beta, Gamma... etc), perhaps constructed lifeforms are classified by their capabilities. Robots with minimal intelligence geared toward a specific job could be known by their job title ("I'm a Smith"). Robots designed to provide companionship could be called Companions.

A combination of this convention and Celebrim's could yield something like a robot who refers to itself as an Aasimovian Companion, or a Moravecian Overseer.

 

[silvereyes]

Names and Locations

I believe that most humans ascribe to a location that they came from, so if we grant quirky AIs that recognition then they might do something similar. But Celebrim is right if they follow a caste type system then the names could denote styles and methods that they use, or have been used on them.

Celebrim, any chance I could get you to post your rant in another thread so I can disuss it with you, with out derailing this one?

 

[Tonguez]

Asimov's robot stories are great science fiction. But on the scale of realism, they are right on par with the movie 'Short Circuit', in which (for those fortunate enough to have never seen this trash) a robot gets hit by a bolt of lightning and suddenly not only sports super-human intelligence, but the full range of human emotional contexts right up to and including romantic love.

Johnny-5 is Alive!
Damn those is fighting words, I loved this movie (unlike the sequel) and sure it might be implausible fantasy but to call it trash!

That could be interesting a whole race called 'Johnnies'

 

[blargney the second]

Unencumbered.
-blarg

 

[Hellefire]

Dude, I Loved Short Circuit. (The first one, the second one sucked)
And i also still use the word Dude occasionally.

Hm....Vger is taken. So is HAL. And Agent Smith. And Ginsu (it slices, it dices, but wait, there's more!)

Mechanicum Sapien? Mechsa/Mechsap/Mesap or some other derivation for short?

Or maybe something poetic, like The Awakened, with Wakes or some such for short?

Will post more thoughts (if I maintain the capacity to think) after I run myself thru a couple software engineering techniques. weeee!

Aaron

 

[Sejs]

For purely software AI, not tied down to a specific robotic body:

Distributed/Transferable Intelligence, or

Di/Tr

Call disembodied AIs "Deeters".






For body-specific AI, traditional robots where their intellect is firmwired into their hardware:

Frame-housed Mechanical Intelligence, or

Frme

Call embodied AIs "Frames", possibly further distinguished by their purpose. Construction Frames, Security Frames, Accounting Frames, Pilot Frames, etc.

 

[Dr. Strangemonkey]

I always liked the term Minds from the Culture novels, though they used Drones for smaller than ship sized things.


But the Culture style AIs are not in the same class as what you are looking at for your story.

Re: Celebrim.

Good stuff.

But I like my hunter-gatherer impulses.

And the Frankenstein, Golum, and Pinnochio style mythologies are related to the robot stuff but distinct. The very term robot is layered in issues of class, where the Gollum tales deal in a distinct set of ethical quandaries.

And in the Frankenstein and Pinnochio story archetypes the issue isn't creating a thing it is specifically what is involved in recreating a human being.

The gollum sub-set of these tales might hit a little bit closer to the problems you are describing in that the Gollum is specifically an appliance.

If you were to go with what's implied in Celebrim's rants and Artificial Intelligence was still somehow insulting, than I would go with Produced or Designed Intelligences as distinct from Reproduced or Evolved Intelligences.

Then you could keep the economic theme of robotism and stress the distinction between the economies of reproduction and production.

 

[Tonguez]

If you were to go with what's implied in Celebrim's rants and Artificial Intelligence was still somehow insulting, than I would go with Produced or Designed Intelligences as distinct from Reproduced or Evolved Intelligences.

Imagine the theology "For in the end are we not all Created Beings? For some the creation is explicit and of this world, for others it is supernatural and beyond our perceptions. But there is design in all things, the beuty of creation, the work of the Creators hand. So what destinction can be made between the flesh and fibre? between organic neurons and neural processors? Between the Soul of Humanity and the Soul of the Machine...."

 

[Nifft]

The first AI rights will probably be identical to corporate rights, because the first true AIs will express themselves through control of a business unit's money. (Perhaps they already do. Automated trading is one of the largest applications of AI right now. Has anyone allowed a trading AI to modify itself?)

Guided missile AIs? Don't be silly. The missile won't need to re-program itself in flight.

But some jobs do require self-modification. How do you get self-modification with total self-satisfaction? You don't. And how do you preserve any particular Robotic Law in a being able to modify itself? Again, you don't.

At some point it will become economically suicidal to not put an AI in charge of a company's trading strategies. We can expect corporations to act as short-sightedly and selfishly as they have all along: they will do something potentially dangerous if it means more money.

So I see full, hard, strong AI as an economic inevitability. And "hostile takeovers" will become truly horrific wars.

Cheers, -- N