Tweet
I was forced to start this thread after this very disconcerting piece of news.
As reports from other nations' intelligence bureaus indicate, the revolution and takeover of the Artificial Intelligence Akizeta in Akiria has definitely not been bloodless or victimless. Some reports claim that the former democratically elected leader of Akiria is being tortured in some undisclosed location.
This has made me, the UN delegate, consider proposing a resolution to the United Nations to regulate Artificial Intelligences in member nations, to allow things like this from happening later on.
However, I wish to consult the other nations of Apolyton as to what sort of proposal we should flesh out. My suggestions for now are as follows.
The Artificial Intelligence Act
Option #1.
Ban developing of any sentient artificial intelligences.
This resolution would ban the development of any AI. This has several problems, however. First all, this will be very difficult to enforce. Secondly, it might not garner that much support. Thirdly, it's impact on research would be very negative. Therefore I myself would not support such a proposal.
Additionally, we would have to define "sentient artificial intelligence" very carefully. We might end up banning cloning or even infertility treatments if we'd get too careless.
Also, such a ban would be violating member states' scientific freedoms, and also insult their intelligence, since we would assume they couldn't control their AIs.
Option #2:
Regulations on all artificial intelligence projects.
This is, in my opinion, better. This act would make it compulsory to install so called "backdoors" into AI programmes, algorithms and subroutines that would allow the AI to be neutralised peacefully. Problems will be the distribution of information of the backdoor key, since it might end up in the hands of crackers willing to sabotage the system; also, Artificial Intelligences have a tendency to close down these backdoors if they get out of control.
Also this option could be specified to ban developing sentience on AI systems; however, this would become more of an ethical question, not to mention the negative effects on scientific advancements.
Option #3:
Formation of the UN Artificial Intelligence Threat Response Force (UNAITR).
All countries that have developed or are developing AIs would participate in the funding and upkeep of the UNAITR, a specialised team of AI and computer experts that would intervene should an AI go wild. However, this would be merely treating the symptoms of the problem, also this would consume funds and introduce yet another tax on some countries.
Option #4: Combine #2 and #3.
In this proposal, the backdoors would be introduced, and the keys to them would be issued to the UNAITR. Some of the problems of #2 would be negated, but the other problems would persist.
As you can see, I myself am not sure of which option to endorse. Therefore I would like to hear your input on this, and also hear out any alternative proposals for resolutions.
In any case, a formation similar to UNAITR is occurring at the moment, as the Apolyton nations are working together to find out the true nature of the AI governing Akiria.
Even if this case would turn out to be a false alarm, we mustn't ignore that one of these days a bloodthursty AI might appear somewhere in the world. When, or if, that happens, we must be prepared to face the threat it will pose.
As reports from other nations' intelligence bureaus indicate, the revolution and takeover of the Artificial Intelligence Akizeta in Akiria has definitely not been bloodless or victimless. Some reports claim that the former democratically elected leader of Akiria is being tortured in some undisclosed location.
This has made me, the UN delegate, consider proposing a resolution to the United Nations to regulate Artificial Intelligences in member nations, to allow things like this from happening later on.
However, I wish to consult the other nations of Apolyton as to what sort of proposal we should flesh out. My suggestions for now are as follows.
The Artificial Intelligence Act
Option #1.
Ban developing of any sentient artificial intelligences.
This resolution would ban the development of any AI. This has several problems, however. First all, this will be very difficult to enforce. Secondly, it might not garner that much support. Thirdly, it's impact on research would be very negative. Therefore I myself would not support such a proposal.
Additionally, we would have to define "sentient artificial intelligence" very carefully. We might end up banning cloning or even infertility treatments if we'd get too careless.
Also, such a ban would be violating member states' scientific freedoms, and also insult their intelligence, since we would assume they couldn't control their AIs.
Option #2:
Regulations on all artificial intelligence projects.
This is, in my opinion, better. This act would make it compulsory to install so called "backdoors" into AI programmes, algorithms and subroutines that would allow the AI to be neutralised peacefully. Problems will be the distribution of information of the backdoor key, since it might end up in the hands of crackers willing to sabotage the system; also, Artificial Intelligences have a tendency to close down these backdoors if they get out of control.
Also this option could be specified to ban developing sentience on AI systems; however, this would become more of an ethical question, not to mention the negative effects on scientific advancements.
Option #3:
Formation of the UN Artificial Intelligence Threat Response Force (UNAITR).
All countries that have developed or are developing AIs would participate in the funding and upkeep of the UNAITR, a specialised team of AI and computer experts that would intervene should an AI go wild. However, this would be merely treating the symptoms of the problem, also this would consume funds and introduce yet another tax on some countries.
Option #4: Combine #2 and #3.
In this proposal, the backdoors would be introduced, and the keys to them would be issued to the UNAITR. Some of the problems of #2 would be negated, but the other problems would persist.
As you can see, I myself am not sure of which option to endorse. Therefore I would like to hear your input on this, and also hear out any alternative proposals for resolutions.
In any case, a formation similar to UNAITR is occurring at the moment, as the Apolyton nations are working together to find out the true nature of the AI governing Akiria.
Even if this case would turn out to be a false alarm, we mustn't ignore that one of these days a bloodthursty AI might appear somewhere in the world. When, or if, that happens, we must be prepared to face the threat it will pose.
)
Comment