Anthropic selection declarethe company selected Accenture as a partner to conduct an independent assessment of its artificial intelligence systems. Accenture experts will be directly involved in the AI Lab workflow. The partnership is the first step in implementing Anthropic CEO Dario Amodei’s initiative to slow the development of artificial intelligence.
Image source: anthropoic.com
According to the agreement, the two companies will invest at least US$1 billion in “capacity building in this field” over the next five years. However, given the “importance and urgency” of the task, Anthropic will provide direct funding for Accenture’s work. “In the longer term, we believe funding should come from collaborative or public sources, as we called for in our vision for advanced artificial intelligence released in June. Since there is currently no such source, we plan to work with various evaluation agencies under different funding conditions” said the human.
Last week, Dario Amodei proposed a three-phase plan to slow the improvement of advanced artificial intelligence models. The first phase involves the involvement of independent experts in laboratory work to monitor compliance with safety requirements, stress test the model and assess its compliance with human values. The initiative is backed by OpenAI CEO Sam Altman and xAI CEO Elon Musk. Senior executives at other technology companies, including Nvidia CEO Jensen Huang, are skeptical of the proposal and believe additional regulation is not necessary.
Anthropic’s collaboration with Accenture is not exclusive: the company is in talks with nonprofit research organization METR and other potential partners. “We are sharing these initial steps so that the public and other AI developers can understand how our work works. We anticipate that our approach will change as this practice evolves. As the project progresses and more evaluators are recruited, we will share new information“The human added.
If you find an error, select it with your mouse and press CTRL+ENTER.
