Well I'm currently thinking of programming a "learning" chat bot. The first idea I gathered was that people might intentionally attempt to pollute its knowledge base. The solution I came up with is to grant people clearance accesses. I will grant myself the highest clearance, and if there are people I trust to act proper when communicating with it, I will give them better clearance. People it barely "knows" will not be able to modify its knowledge base, but only to communicate with it.

Cool, let me know how it goes if you will...would be...revolutionary if it learns and speaks English and stuff. What sorts of learning and things will it do?

Quote:

Original post by Max_Payne Well I'm currently thinking of programming a "learning" chat bot. The first idea I gathered was that people might intentionally attempt to pollute its knowledge base. The solution I came up with is to grant people clearance accesses. I will grant myself the highest clearance, and if there are people I trust to act proper when communicating with it, I will give them better clearance. People it barely "knows" will not be able to modify its knowledge base, but only to communicate with it.

IRC is really the worst place to develop a learning AI. You will wind up with a really psychotic robot who's knowledge base is filled with garbage.

There are no intelligent lifeforms on IRC to learn from.

Quote:

Original post by Name_Unknown Quote: Original post by Max_Payne Well I'm currently thinking of programming a "learning" chat bot. The first idea I gathered was that people might intentionally attempt to pollute its knowledge base. The solution I came up with is to grant people clearance accesses. I will grant myself the highest clearance, and if there are people I trust to act proper when communicating with it, I will give them better clearance. People it barely "knows" will not be able to modify its knowledge base, but only to communicate with it. IRC is really the worst place to develop a learning AI. You will wind up with a really psychotic robot who's knowledge base is filled with garbage. There are no intelligent lifeforms on IRC to learn from.

Well, there are also some intelligent people on IRC ;) The point is to have an interface with it, and a way to show it to other people.

Quote:

Original post by Max_Payne Well, there are also some intelligent people on IRC ;) The point is to have an interface with it, and a way to show it to other people.

I guess they are hiding themselves in private channels ;-)

Really, I won't discourage you, but all your ideas I tried years ago and many more you haven't even thought of yet. The medium is just.. not a good reflection of normal human social interaction ;-)

But, if you do, anchoring is a good idea, give it some clues about trustworthiness . The first goal is to get it to survive on IRC. If someone realizes it is a bot they will definitely attack it.

I'd like to encounter your "being", who seems like a social misfit. Is it around?

Quote:

Original post by verdad All things are consciousness.

This is an opinion, not a theory or a fact. Prove it.

Hmmm. My new resoning engine currently allows people to add links and rules, with o(n) or even o(log2 n) complexity, where n is the number of links, in the superlink.

From,
Nice coder

Quote:

Original post by verdad I'd like to encounter your "being", who seems like a social misfit. Is it around?

Were you talking to me?

If you have been on IRC for a few years back you probably encountered at least one of my 'beings' ... my goals were to make it hard to recognize it was a robot, whether it was AI, I can't say, but I have certainly seen people chatting with them thinking they were real people ;-)

IRC is a hostile environment on many levels. You obviously haven't written bots very long if you haven't noticed this. If you stick to a safe channel, ok... but if you code bots with an instinct to explore, you will see hostility on many levels.

Servers can get annoyed by bots who don't play by the rules. To the extreme, a badly behaving bot can get you k-lined. Some servers don't even allow bots at all.

Many people think bots == hackers. Not true, none of my bots were designed to take over channels, etc. I had different prototypes, explorers, learners, observers, collectors.

People can get annoyed by bots if they recognize robotic behavior.

People can be down right aggressive if they recognize a bot in some channels.

Some channels should be avoided period. An explorer bot might not realize this! Which means having to give it a list of rules on which channels NEVER to go into..

is it possible to add new custom node to a layer of a neural network at runtime

so that the computer generated the use and the operations of this node on its own?