Reading 2 minutes Views 5 Published Updated
OpenAI, the parent company of AI chatbot ChatGPT, has launched an initiative to contribute more democratically to AI development.
In a May 25 official statement, the company said it was preparing to award 10 grants of $100,000 each for experiments to create a “proof of concept,” a democratic process for defining the rules that AI systems should follow.
According to OpenAI, the rules must be “within the limits determined by law” and must benefit humanity.
“This grant represents a step towards establishing democratic processes for the oversight of AGI and, ultimately, superintelligence.”
The company said the experiments will form the basis for a more “global” and “ambitious” project in the future. He also noted that the conclusions of the experiments will not be binding, but will be used to explore important issues related to the management of AI.
The grant is provided by the non-profit arm of OpenAI. According to him, the results of the project will be free and publicly available.
Related: OpenAI Launches Official ChatGPT App for iOS, Android Coming ‘Coming Soon’
This is because governments around the world are looking to implement rules on general purpose generative AI. OpenAI CEO Sam Altman recently met with regulators in Europe to highlight the importance of non-restrictive rules and not hinder innovation.
A week earlier, Altman had delivered a similar message to the US Congress.
In the announcement of a new grant program, OpenAI reiterates the view that laws must be adapted to technology and that AI needs “more complex and adaptive rules for its behavior.”
It gave examples of questions such as: “How should controversial views be represented in AI results?” After which it was said that no person, company or country should dictate such decisions.
OpenAI has previously warned that if AI is not developed carefully, a superhuman form of AI could emerge within a decade. Therefore, the developers “must do it right”.