Update: September 18, 1996

Now, let's face what differentiates the truly intelligent human person from the machine: responsibility.

A machine, today, acts without the judgement of its previous actions. A machine is not able to look back and consider if what it has done was good or bad, nor is it able to consider if its planned actions entail some possible harmful consequences.

Responsibility has something to do with feedback but at a higher level.

Feedback is just mechanical while responsibility is touching the whole person. Responsibility has to do with the core circle of the mind, it's one of its highest function. A shortage of it can be disastrous for a complex system.

What is the word which has the most meaning in your life? It is your own surname. How do you feel when you pronounce it: happy, unhappy, indifferent, troubled? You know it, it's your own problem, it's your responsibility, it's not mine. It's a mirror of yourself.

This is to explain that a name or a surname would be for an I-Agent the starting point of a feeling of responsibility.

We can imagine a lot of functions which could be able to modify the status of this variable called surname. They could modify it's weight, giving it a positive or negative feedback. If this feedback is positive: "OK, go ahead". If it's negative: "Oh oh, there's a problem. Something must be done". This could trigger some actions.

Now, responsibility on the individual level is not always sufficient. When it becomes uncontrollable, society has to react. That's what we call Justice.

Although, today, it's probably too early to think about justice applied to I-Agents, we can imagine another type of feedback or safeguard given by society: some independent organisms could give a rating to individual I-Agents depending on their behavior observed in cyberspace.

Each I-Agent should be able to indicate to any person or any other I-Agent its name and identity. The person or the I-Agent receiving a request from an I-Agent could then verify its rating and then react accordingly.

We do the same, when we are meeting a person and that we are considering his or her reputation.

Of course, I-Agents could then be tempted to hide their bad actions like some humans would do.

But that's another story. I've never pretended that I-Agents would be better or worse than we are.

Return to Main Page