Microsoft CTO Kevin Scott on how AI can save the web, not destroy it
May 19, 2025
auto_awesome
Kevin Scott, Microsoft’s Chief Technology Officer, discusses the future of search and AI's transformative role. He shares insights on Microsoft's open-source tool, which enhances website interactivity through natural language search. Scott delves into the concept of the 'agentic web,' emphasizing the importance of user autonomy amid AI integration. He also explores how AI can impact creative industries, addressing concerns about authorship and copyright while advocating for a more equitable web for creators and consumers.
Kevin Scott discusses Microsoft's open-source AI tool that allows websites to integrate natural language search, enhancing local search capabilities.
The Model Context Protocol is crucial for creating a decentralized web, enabling structured AI interactions and reducing reliance on centralized search engines.
The podcast highlights the importance of balancing AI advancement with creator rights, emphasizing fair compensation models to sustain innovation in creative fields.
Deep dives
Transforming Search with AI
The discussion centers on Microsoft's innovative approach to revolutionizing local search capabilities by integrating AI-powered natural language search tools into websites. This strategy allows various web platforms to implement natural language search easily and cost-effectively, enhancing user experience significantly. For instance, a demo involving TripAdvisor showcased how quickly their team could use the tool to improve their own search features, highlighting the practicality of this technology. Ultimately, the objective is to establish a decentralized system where websites can manage their own search functionalities, potentially diminishing reliance on large centralized search engines.
The Model Context Protocol (MCP)
The Model Context Protocol (MCP) is introduced as a foundational element for the future of search and web interaction, enabling AI to access databases in a structured way. By allowing agents to interact with different services seamlessly, MCP promotes broader accessibility of data and reduces the need for centralized search indexes. Its potential to create competitive opportunities in the market is significant, as it facilitates a new standard for websites that can ultimately transform the landscape of online search. As more companies begin to adopt MCP, this could lead to more user engagement and varied web traffic, reinventing how content is consumed.
Challenges in the Agentic Era
The transition to an agentic web brings several complexities, particularly regarding business models and the incentives for service providers. As websites adopt standards like MCP, there is a pressing need to determine how businesses will monetize their services within this new framework. Questions arise around whether agents should have unrestricted access to services, and how content creators can maintain agency over their offerings while ensuring profitability. This balancing act is crucial for the long-term sustainability of the web, and it remains a daunting hurdle as various parties navigate this evolving landscape.
AI’s Impact on Creativity
The conversation also delves into the intricate relationship between AI and creativity, particularly regarding copyright and the compensation of creators. With AI's advancing capabilities, there are ongoing concerns about how content is utilized, especially in terms of training data. The implications for artists and creators highlight the need for fair compensation models that respect intellectual property while fostering innovation. As AI continues to transform creative processes, maintaining a dialogue about the rights of creators and the ethical use of their works will be essential.
Future Prospects for AI and the Web
The overall discussion underscores the need for an architectural shift in the web, focusing on open standards and protocols that can enhance collaboration and interoperability across various platforms. By encouraging a more agent-centric environment where user intent drives the interactions, there is a potential for innovation that benefits both users and content providers. As technologies like MCP and NLWeb gain traction, the hope is for a more inclusive web that aligns the interests of creators, users, and platforms. This visionary approach aims to redefine the internet's functionality while ensuring that all stakeholders have a voice in determining its future.
Today, I’m talking with Kevin Scott, the chief technology officer of Microsoft, and one of the company’s AI leaders. Kevin is one of my favorite repeat Decoder guests, and he joined the show this time to talk about the future of search.
Microsoft just announced an open-source tool for websites to integrate AI powered natural language search with just a little bit of effort, in a way that lets them actually run whatever models they want and keep control of their data. I saw some demos before Kevin and I chatted, and the improvements over the bad local search on most sites was obvious. So we talked about what this will mean for AI, for search engines, and for the future of the web.
Links:
Microsoft’s plan to fix the web: letting every website run AI search for cheap | Verge
Microsoft Build 2025: news and announcements from the developer conference | Verge
Introducing the Model Context Protocol | Anthropic
Copyright Office head fired after reporting AI training isn’t always fair use | Ars Technica
Microsoft CTO Kevin Scott on how AI and art will coexist in the future | Decoder
Microsoft CTO Kevin Scott thinks Sydney might make a comeback | Decoder
Microsoft’s CTO explains how AI can help health care in the US right now | Vergecast
