Time for some Friday Big Ideas.
The term “context engineering” is gaining recognition as a better alternative to “prompt engineering” in artificial intelligence. Supporters claim it more accurately describes the skill of providing necessary context for large language models to work well. Well-known industry figures, such as Shopify CEO Tobi Lutke and AI researcher Andrej Karpathy, highlight that context engineering involves a detailed understanding of task descriptions, few-shot examples, and relevant data. This method is viewed as crucial for improving AI application performance because it combines technical skills and an intuitive grasp of how these models function. The change in terminology reflects a wider trend in AI development that focuses on clear communication and the complexity of completing tasks.
A recent article explores the unexpected versatility of the Model Context Protocol, or MCP, highlighting its potential as a universal plugin system. While initially designed to enhance artificial intelligence capabilities, the protocol can connect various tools and data sources, enabling applications to leverage functionalities without direct integration. The author draws parallels to the evolution of car cigarette lighters, which have transitioned from a niche use to a universal power source. As more developers create MCP servers for their AI systems, the resulting ecosystem allows applications to access a wider range of functionalities, akin to a communal potluck of software capabilities. The article emphasizes that this spontaneous development indicates a shift towards a more interconnected technological landscape, where protocols like MCP facilitate seamless interactions between diverse systems.
How about the flood of AI notetakers at meetings. The Washington Post has you covered, AI note-takers are increasingly dominating Zoom calls, as many employees choose to skip meetings in favor of automated solutions. In a recent meeting attended by Clifton Sellers, he observed that six human participants were outnumbered by ten AI note-taking apps, highlighting a growing trend where technology is replacing traditional human interaction. This shift raises concerns about the impact of artificial intelligence on workplace dynamics. While AI note-takers can efficiently record and summarize discussions, their prevalence may hinder meaningful communication among team members. As companies continue to adopt these tools, it is essential to consider the balance between technological efficiency and human connection in virtual work environments.
Why do we care?
Some questions to consider.
Are your teams equipped to think in terms of context, not just commands, when working with AI tools?
If MCP and similar protocols turn AI systems into universal connectors, how does that change your integration strategy for client systems?
Are you prepared to manage the security and governance implications when third-party tools freely interact via protocols like MCP?
Could MSPs leverage this interoperability trend to create their own micro-ecosystems of tools for clients—reducing vendor sprawl and improving stickiness?
What happens to organizational culture when AI tools replace actual human presence in client meetings?
Are you advising clients on AI governance policies that address where automated tools like notetakers fit into workflows (and where they shouldn’t)?