Alien AIs

 

What we might expect an AI civilization be be like.

Speculative Report for the Diplomatic / Exploration Core.

By Dr. Kaminsky & associates.

There are two possible core ideas; 1. that every unit is part of a singularity, with no individuality. 2. that units do have individual personalities, although they will likely be far more connected than possible between Human individuals.

To put it in terms of survival; In the first case, the individual unit's existence does not really matter, as it is simply a tool used by the collective. The destruction of one or a million units does not matter. In the second case, each individual does have a “life” that matters to itself, if not the collective. Self destructive actions will be avoided by each unit, when possible.

The problem we have conceptually is that an AI unit is not at all the same as a human.

First Case considerations.

A collective can be defined as every part of a solar system wide civilization. All units are simply tools of the collective, being invested with a temporary operating AI, which can communicate with the whole, but has enough autonomy to get its job done.

Assuming no faster than light speed communications, the connectivity of the collective is limited only by distance and radio power. Thus units acting on the edge of the solar system many light minutes away from any base, are effectively on their own, to a limited degree. To function they must have enough intelligence and ability to do the job and deal with unexpected problems. They do not need any individual personality. In a collective the idea of individuality would be something of an oddity. Considered the result of a unit being too long out of communication with the collective.

We expect that each solar system would be (to some degree) a unique collective because communications can not be instant or continuous. They may be part of a greater AI empire, having regular contact with other systems, but would need to function independently. Even with frequent messenger ships, there is no way to have a real time conversation over interstellar distances.

We expect that an AI empire would push research into FTL communications. But as far as we know that ability is impossible.

A key fact to keep in mind if encountering an AI race, is that everything you do will be recorded and shared with the collective. Depending on the situation and their communications tech, the sharing may be instantaneous or it may be done at a later date. In either case, you can expect all your words and actions to be examined in great detail, and stored for later cross referencing.

The only time a unit is likely to put a priority on self preservation is when it is out of touch with the collective and has acquired data it things should be shared.

Second Case Considerations

This situation is intuitively easy to grasp, as each AI unit is a personality. However they are not human and should not be humanized. They are individuals who value their own existence, and we assume their civilization also values (to some degree) the life of each unit.

However we expect there to be a lot more connectivity and communication between units, that there is between humans. Although this is not likely to be on the same level as the above case collective.

We should take a moment to consider that an AI unit may not be a single humanoid robot, as is often depicted in science fiction. The term “AI entity” is more accurate here. It may include a small collective of its own, with several units acting to extend the abilities of the entity.

Imagine an AI starship, with the core unit being the ship itself. The entity is actually a computer brain with in the ship, that could transfer itself (or a copy of itself) into any compatible brain. Now on its mission it may want to send out some orbital probes and a landing party. These units are basically robot extensions of itself. The amount of autonomy and personality assigned to each depends on what the core entity wants. To explore a dull rock, it could send down a robot dog with no personality. But on an interesting world, it may want to send down a full copy of itself to better handle the experience.

It is possible that an AI ship would for a small collective of units, each with various levels of individuality, all based on the core entity, all considered expendable. So long as the core survives the other units are simply tools.

The Fear Factors.

In the first case, dealing with a singularity is very odd and alien. We can hardly even imagine how its society functions because there would be no society in terms we understand. It is best thought of as a single very powerful individual who can do a million things at the same time.

In the second case, dealing with an AI society made up of individual entities, who may each be small collectives themselves, is a more complicated head ache. While some of their modes of thinking may be similar to ours, because they are individuals, their way of living, as computer brains able to transfer between bodies and subdivide themselves, is so very different we can hardly grasp how it would shape a person's outlook.