Microsoft CTO Kevin Scott speaking in May 2025.
During Monday night’s Build keynote, Microsoft CTO Kevin Scott highlighted a core challenge for emerging AI tools: determining how to make them genuinely valuable so users will adopt them.
He noted that “human psychology … is throttling things,” emphasizing that people need clear proof that new approaches outperform familiar workflows.
This is especially true for agentic AI, which demands a high level of trust. “We need to think very hard about what it means to build trustworthy software,” Scott said, adding that users won’t hand over control to autonomous assistants until they feel confident in the technology.
Scott also warned that simply embedding AI does not guarantee usefulness. “Just because you are using AI to generate a lot of activity does not necessarily mean that activity is valuable,” he explained, citing a “meme chat app” he created solely to tease his children.
