In March of last year, the Future of Life Institute published an open letter. The aforementioned letter demands that AI labs immediately halt all activities related to the development of artificial intelligence systems more potent than GPT-4 for at least six months. Additionally, they decided that the aforementioned pause needed to involve all important actors and be visible and verifiable.
They further stated that governments should intervene and impose a moratorium if businesses refused to comply with the aforementioned stop. Furthermore, the aforementioned letter added that sophisticated artificial intelligence systems should only be created when researchers are certain that their benefits would outweigh their risks.
More than a thousand people, including many eminent figures in technology, originally signed the aforementioned letter. As soon as the letter is published, additional signatures start to come in.
Reason 01: Delayed AI Benefits
Numerous sources have previously demonstrated how artificial intelligence will improve many other fields, including education, medication research, medical diagnosis, climate modeling, and weather forecasting. On the other hand, the recent letter that demands a government-mandated halt to work on a crucial area of AI will inevitably postpone access to the advantages of the technology.
The employment of massive language models to enhance education and increase access to legal services may be reflective of these delays or adverse impacts, and those outcomes could be problematic.
Reason 02: Legal Uncertainty
It is of little consequence that neither the federal nor state governments of the United States are in a position to impose a moratorium on the development of sophisticated artificial intelligence systems. Protecting the public from deceptive or unfair business practices and from unfair methods of competition” is the Federal Trade Commission’s (FTC) stated objective.
Although Congress has broad legislative authority under the Commerce Clause, that authority also has limits. What’s interesting in this situation is that the FTC moratorium is intended to prevent companies from competing to develop better artificial intelligence systems, which would force them instead to act in lockstep to stop their work on training large artificial intelligence models. To cut to the chase, the moratorium would be challenged in court without regard to whether it was imposed by a government agency or by Congress.
Reason 03: Difficulty in Implementation
A moratorium would be challenging to successfully enforce in addition to those other factors. As the American government won’t begin conducting raids on businesses believed to be practicing illegal artificial intelligence training in the manner of the Prohibition era.
This essentially meant that the government lacked the personnel and technical means necessary to formally confirm that a national moratorium was being adhered to. Given this, a moratorium would probably be enacted through a self-reporting procedure that required businesses and academic institutions to declare that they were not performing illegal artificial intelligence work.