The article states (in section 5) that “regulators controlling the use of generic computing hardware and data storage may be more important to determining the future of artificial intelligence than those that design algorithms”
How will this work exactly? Any regulatory attempt to prevent acquisition of enough computing power to create an AI would likely also make it difficult to acquire the computing and storage equipment one might need for a more modest purpose, (e.g. a server farm for an online game or hosted application, the needs of a corporation’s IT department, etc.). Government regulation of computing equipment sounds almost as dystopian as the scenario it is intended to prevent.
And, while a government could (perhaps) prevent its citizens from creating AGI superintelligence this way, how would it prevent other governments and entities not under its jurisdiction from doing so? Gone are the days when any one nation has a monopoly on high-end computing resources.
The article states (in section 5) that “regulators controlling the use of generic computing hardware and data storage may be more important to determining the future of artificial intelligence than those that design algorithms”
How will this work exactly? Any regulatory attempt to prevent acquisition of enough computing power to create an AI would likely also make it difficult to acquire the computing and storage equipment one might need for a more modest purpose, (e.g. a server farm for an online game or hosted application, the needs of a corporation’s IT department, etc.). Government regulation of computing equipment sounds almost as dystopian as the scenario it is intended to prevent.
And, while a government could (perhaps) prevent its citizens from creating AGI superintelligence this way, how would it prevent other governments and entities not under its jurisdiction from doing so? Gone are the days when any one nation has a monopoly on high-end computing resources.