Tool-based works might be a faster and safer way to create useful AI, but as long as agent-based methods are possible it seems extremely important to me to work on verifying friendliness of artificial agents.
Important, perhaps, but extremely important? If tool-based systems are faster in coming and safer, then they will be available to help the process of creating, studying, and (if necessary) defending against powerful agents.
My prediction would be that tool AI would be economically incentivised, since humans want tools. Agent AI might be created later on more aesthetic grounds, as pets or hoped-for equals. (But that’s just an intuition.)
For the same reason that a personal assistant is vastly more useful and powerful than a PDA, even though they might nominally serve the same function of remembering phone numbers, appointments, etc. people are extremely likely to want to create agent AIs.
Tool-based works might be a faster and safer way to create useful AI, but as long as agent-based methods are possible it seems extremely important to me to work on verifying friendliness of artificial agents.
Important, perhaps, but extremely important? If tool-based systems are faster in coming and safer, then they will be available to help the process of creating, studying, and (if necessary) defending against powerful agents.
My prediction would be that tool AI would be economically incentivised, since humans want tools. Agent AI might be created later on more aesthetic grounds, as pets or hoped-for equals. (But that’s just an intuition.)
For the same reason that a personal assistant is vastly more useful and powerful than a PDA, even though they might nominally serve the same function of remembering phone numbers, appointments, etc. people are extremely likely to want to create agent AIs.