Refusing agents or use cases


#1

Is there any way how to refuse / block / delete an agent or forbid an activity of an agent? To prevent adding military or other malevolent agents or to prevent using otherwise neutral agents (eg. facial recognition agent) for malevolent purposes?

Just to avoid scenario when SNET is used in aplications like what you can see in the video below. Everyone will have access to the platform so it would be very easy for eg. terrorist groups to benefit from it.


#2

Architecturally no. When you create a decentralised network you necessarily give away a lot of control.

However, in terms of our official registry and reputation ranking, we will have the ability to control what services are listed and which one’s are endorsed by the SingularityNET foundation. At least initially, this will be important for people to have some faith that the services they pay for will work effectively. Those that don’t will quickly be ranked badly and removed from our registry.


#3

mathematics was used on creation nuclear bombs, should we forbid mathematics?


#4

So it is possible we will have a killer agents in the network, benefiting from the network and no means to stop it?


#5

Well, nobody is able to create a nuclear bomb with just mathematics. You still need some very scarce materials etc. so building such a bomb is very difficult even for governments.

With technology, it will be increasingly easier for small groups or even individuals to do a big harm. It is not that difficult to build a few drones, pay a few bucks for outsourcing AI if it is available in the marketplace.

Not sure we as society will be able to survive without carefully watching each other. I value liberty a lot so I don’t like it but I am not sure it will be possible.


#7

as i know Snet’s going to be decentralized sysem, isn’t it? so it’s not good idea. May be voting on listing/delisting AI servises would be good desicion


#9

That’s what I wanted to suggest. You were faster and I owe you a beer :slight_smile:

We will have a democratic mechanism anyway so why not having an option to vote for shutting an agent down. As it is almost always a minority of people who have malicious intentions.


#10

so what’s about AI providers? how they can be confident in marketplace? such censorship unhealthy influence on market


#11

How I feel it, it is not a censorship but a security measure. The community would have no interest in shutting down agents that are not harmful.

And we won’t shut down neutral agents (eg. face recognition agents) just because they were used by malicious ones.