This article explores the potential of large language models (LLMs) to shift computing away from application-centric models towards a more dynamic, relational, and human-centered paradigm. It argues that LLMs can unlock the computer's full capabilities by mediating human intent and enabling a conversational style of interaction.
This paper introduces Pointer Assistant, an LLM-powered chatbot designed as a visual pointer alongside the user's mouse cursor to proactively assist with on-screen tasks. A study with 220 participants performing a financial budget planning task showed that the pointer-based, proactive design reduced task load, improved satisfaction, and increased the number of ideas generated compared to a traditional chat log interface. While participants found the assistant fun and helpful, feedback suggests improvements to its assertiveness could further enhance the user experience of human-AI collaboration.
This blog post introduces the Semantic Telemetry project at Microsoft Research, which uses a data science approach to analyze how people interact with AI systems, specifically focusing on Copilot in Bing usage. It discusses the complexity of human-AI interactions and how they differ from traditional search.
- Topics: Copilot in Bing chats were analyzed for topic categorization. Technology (21%) was the most common topic, followed by Entertainment (12.8%), Health (11%), and others. Within technology, programming and scripting were prominent subtopics.
- Platform Differences: Mobile users tend to use Copilot for personal tasks, while desktop users engage in more professional activities.
- Topics: Copilot in Bing chats were analyzed for topic categorization. Technology (21%) was the most common topic, followed by Entertainment (12.8%), Health (11%), and others. Within technology, programming and scripting were prominent subtopics.
- Platform Differences: Mobile users tend to use Copilot for personal tasks, while desktop users engage in more professional activities.
A new model developed by researchers at MIT and the University of Washington predicts human goals or actions more accurately than previous models. The latent inference budget model identifies patterns in human or machine decision-making and uses this information to forecast behavior.
