Satya Nadella, CEO of Microsoft, addressed the growing issue of AI safety in an internal memo, stressing the importance of human control over this technology. He called for more extensive testing by companies operating in this field and warned that companies must take the necessary time to secure this technology.
The Need for Human Control Over AI
Nadella noted in a memo published earlier this week on a company message board that with increasing risks, more time must be spent ensuring the safety of the technology. He stated, "If you do not take the necessary time, you will ultimately lose your license to operate." These remarks came in response to serious warnings from some AI researchers who stated last week that this technology could pose a threat to humanity.
Read more: Nvidia is valued lower with Piper's forecasts
Responsible AI Development
In this context, Dario Amodei, CEO of Anthropic, discussed the need to slow down AI development and announced that his company would allow external oversight of its model progress. Additionally, Sam Altman, CEO of OpenAI, emphasized the importance of third-party oversight and stated that under necessary conditions, the deployment of technology should be delayed to ensure that it is safe and compliant.
Microsoft also published its 37-page draft ethical code for "human-centered" AI models on Monday. Mustafa Suleyman, CEO of Microsoft AI, said, "The concerning issues we have long theorized about are becoming a reality." He also stressed the need to create a boundary ecosystem where closed and open-source models can coexist and added that more care should be taken regarding the timing of AI compliance.
Suleyman further emphasized, "We must ensure that our customers maintain control over their unique and implicit knowledge so they can build their continuous learning machine without relying on any model provider. This is truly what Foundry enables."
Emphasis on Collaboration with Local Communities
He noted that Microsoft must obtain permission to build the infrastructures that empower this technology, ensuring that these benefits are felt locally. The company's data center in Quincy serves as a good example of simultaneous growth with local communities. Suleyman, while referencing the company's real actions to achieve these goals, stressed the importance of creating an integrated approach in three areas: how to build technology at the frontier, how to empower customers to maintain control over their knowledge, and how to ensure the economic benefits are distributed across various communities and countries.
Ultimately, by releasing this memo, Microsoft aims to clarify its approaches to AI and emphasize the importance of safety and human control in this technology.
Read more: Strengthening the AI position of Advanced Micro Devices with the Halo station · Asian indices decline due to concerns over AI slowdown and rising oil prices
