Microsoft: "People matter more than AI"
Microsoft is taking a harder line than rivals on AI design.
TL;DR
- Microsoft's AI chief Mustafa Suleyman advocates for maintaining human control over AI development.
- The company has released a "Humanist AI" code of conduct prioritizing human agency and control.
- Microsoft is willing to compromise on advanced capabilities to ensure AI remains subordinate to humans.
- AI should be designed as a tool to increase human capability, not foster dependence.
- The code rejects the pursuit of "all-purpose superintelligence" that could exceed human control.
- Microsoft plans to revise the code after a public consultation and implement it for model development starting in 2027.
- The company believes AI should not imitate consciousness to ensure it remains controllable.