Well..

You see, I'm running a roleplay for my friends on a small forum. It's set in the future, with a rather low-realism theme (FTL Travel, Psychic Powers, ect.)

I'm trying to make a nice world/universe for them to wander around in.

While I've got a good bit of the solar system (Since we're starting near Earth) about how I want it. The overbearing and corrupt Terran Empire, which controls a large portion of the planet. The fractious smaller nations and states that get bossed around by the empire.. The off-world colonies on the moon, mars and venus (Of which, Mars is a bit of a slum, the Moon Cities are stuck up snobs and the people on Venus tend to be parochial and unfriendly.).. There's also a few spinning ring-cities that orbit the various planets, and the free-floating mining companies (Mostly automated) and various private trading businesses.

While I've got humans more or less down pat, I run into a few issues with AIs. Primarily, what kind of jobs would they excel in.

They aren't much smarter than humans, but have a few mental advantages.

They have perfect memories and can access stored information on networks and the internet.

Some of them can do true multitasking, taking on several tasks at once.

Some of them can think much faster and more thoroughly than a human.

All of them are software. They can transfer themselves from one robotic body to another with some degree of ease.

Now, what kinds of jobs would they be skilled at?

I'd imagine that they would excel at any science/math related positions versus humans and be poor at anything physical or manual.

Generally speaking, humans are good at "discrimination" (in the scientific term) and "classification". That is they are good at grouping objects into different categories. Computers are typically much better at brute force computation, exact arithmetic and repetitive tasks. So AI would be better "boring" tax accountants, auditors, risk management and so on. They'd probably make good programmers, hardware designers, long-term city planners, project cost estimations (ie. estimators in various industries).

Don't forget archivist. Right now computer programs are researching case law and data mining all kinds of large data bases.

There would be weapon systems, transports, powerplants etc. where AI's would do some watch standing. They never tire, can compare data from days, weeks or years past and could even be programed to chit-chat with their human counterparts/co-workers.

Pilots, probably, for air, ground and space craft. I've always thought human pilots would become pretty redundant after AI, if it's the type that can perform flawless physical calculations while simultaneously monitoring all the craft's systems directly and seeing in a range of spectrums.

And yeah, more generally, I think they'd be good at any job that involved machine operation because they could incorporate the machine, whether it's a building, craft or utility system, actually into their own nervous system and directly experience what's going on.

Actually it might be easier to ask which jobs they might lose to humans in. I suppose they might lack social skills due to a radically different and potentially incompatible perception of reality. But after a point they might be able to compensate for that through analysis, the same way a psychopath does.

edited 5th May '12 3:40:21 AM by Kesteven

Everything.

Dangerous jobs. Things that need to be done in outer space, frigid planetoids, places that a human would need a lot of time and equipment just to survive a robot would function just fine in.

Ai could run a complex automated factory, control a multitude of worker drones all at once, advanced computations and calculations for science, maintain an adjustable satelite netowrk , direct ground, sky, and space traffic all at once etc.

An android brain could possibly be the only intelligence in the setting capable of plotting FTL jumps or operating some other kind of high-energy, extremely sensitive and dangerous phlebotinum, forcing their integration into human society in vital jobs having to do with dangerous but highly advanced technology.

Maybe artificial intelligences are capable of surviving FTL travel, whereas FTL travel liquifies organic brains. Or maybe the humans believe there are certain jobs that are better left to machines, for cultural reasons.

Judge, Jury and Executioner

edited 13th May '12 6:35:42 PM by fulltimeD

Define "smarter" than a human. Computers can already compute faster than humans by the millions. Can they think like humans, and yet are still smarter? If they can, well, you can't have them doing "cold logic" jobs because they are so much better at it than humans, because in order for them to think like humans, they're going to have to make mistakes. Unless the software is about deductive reasoning and has nothing to do with the human brain, but can act "like a human". Then detective work might be a good job. As they are robots, hazardous jobs are safe too.

Computers would be able to do just about any job, really. Their one limitation would be lateral, "outside-the-box" thinking. Humans also have an advantage in being more capable of acting unpredictably. A computer can fly a standard plane better, but humans would make better fighter pilots.

I'd be cautious about generalizing here - that's certainly the trope, but the truth is we have no idea what the strengths and weaknesses of computer systems will be like by the time we have software entities advanced enough to be called A.I.

Not to mention that in terms of 'acting unpredictably', humans have all kinds of proven subconscious biases in 'random' generation, as illusionists well know. It probably wouldn't be that hard for a specialised advanced computer system with enough information to predict to within an acceptable range or error what a human will do, while a computer could use all kinds of complicated rotations of algorithms on almost any seed.

edited 27th May '12 7:53:24 AM by Kesteven

Thanks everyone.

I think I have enough ideas to go on now.