The Grand Assembly,
Noting the that many nations are on the brink of developing advanced artificial intelligence systems
And clear that there is not a guarantee that any and all possible danger has subsided or will subside, justifying natural concern about the implications of this new technology,
Though hopeful that nations and civilization will traverse down the path of peaceful existence alongside these new developments, that it is clear that dangers have not been alleviated and removed,
And so, amidst our worry, we hope to avoid a tragedy, one that is preventable and requires swift and precise action to deter and contain the dangers of artificial intelligence, including both on an individual basis and a collective basis and hopes to avoid the doom scenario of an information singularity, an ever-present danger that there is no definitive proof is gone,
Addressing the incorrectness, flawed assumptions and details of GAR #354, which presents many errors and shortcomings in its mission, those flaws being:
As per 1.
The definition of Artificial Intelligence, or AI, is as above described as being able to have ''moral reasoning'', a term so broad and diverse, that lacks a consensus on its definition, assuming it actually exists. As such this should not be a way to measure intelligence, for this is incredibly open to interpretation on its character and existence. It is also not clear when an AI would leap from making purely logic-driven decisions to ''moral reasoning'' if it indeed exists. And there would be difficulty in verifying this was accomplished on its own without bias of input of behalf of its creator and environment,
And how in 2.
Finds a contradicting in giving equal rights to machines while making exceptions purely for apparently convenient purposes;
And that it fails to state why the an advanced system should be garnered rights simply for being overly successful as a result of technological innovation to the point it can achieve human-level mimicry;
And while commending the efforts in 3.
Which was intended to stop the ever-present risk of a ''gray goo'' scenario, 3. however fails to include an additional provision preventing machines classified as non-intelligent involved in the practice of autonomous self-replication from evolving and learning enough to reach a point where it presents a tangible threat of such a scenario;
And while 6.
Notes how while it allows nations to prepare for possible uprisings and disruption from malignant AI, it does not even attempt to implement a basic standards on nations to have any preparation for such a scenario, and additionally finds it contradictive, as in the opening statements in the documentation, it states how the danger of such a scenario has been conquered and avoided, yet in its encouragement of nations to voluntarily prepare for a disaster is completely illogical and hints at the possibility that such danger may still be ever-present despite the current peace enjoyed by nations,
And so with the original documentation filled in error, and a failure to implement and enforce basic safety standards and provisions for dealing with advanced artificial intelligence, putting the lives of citizens in mortal danger and with undetermined risk,
And hopeful that a more suitable replacement, one that secures and enshrines basic standards on dealing with AI will be made in the near future, and with this;
the World Assembly Hereby Repeals General Assembly Resolution #354, Artificial Intelligence Protocol.
