In a development certain to offer huge relief to a public fatigued by political gridlock, Google has agreed to replace the entire U.S. Senate with a simplified version of its AlphaGo artificial intelligence. Voters in all 50 states quickly approved the ballot measure to replace their human elected officials with Google’s technology.
Google will be simplifying the system that it trained to play Go for its new role as the U.S. Senate. A person close to the development of the newly modified system explained that training the AI for political service will take much less time than training it to play Go.
“It turns out successfully executing the duties of political office, at least in the US Congress, does not require anything like true general artificial intelligence that can pass for human. It is sufficient to simply teach a narrow AI system such as AlpaGo the basic requirements of the job.”
Additionally, the system will use considerably less processing power, and employ only moderately sophisticated deep neural networks. It appears most of the current decisions facing the Senate are easily deliberated by a straightforward expert system trained using historical sessions when legislation was actually passed. To draft new legislation the system will aggregate and understand the will of the people by actually asking them via voice, chat, and website forum.
Initially each state was planning to send its own individual AI system to the Senate chamber, however it soon became clear that a single system would be more cost effective and could still easily offer a record of debate to play back for broad media consumption.
Legislation to replace the House with a similar system is under debate in many states, although the broad expectation is that it will be at least another year until every state has signed on. Leading economists agree that until the House follows the lead of the Senate we can expect markets to stagnate, and unemployment, or at least under-employment, to remain relatively high.
There remains a minority of the population concerned that this development represents the beginning of the end, that once the machines are in power humans will ultimately suffer a dire fate. However the majority agree that although an AI apocalypse is a minor concern, it is abundantly clear that electing a bunch of human representatives no longer works, and this is a far better and more trusted alternative.