A Human and an AI Agent Walk into a Bar...

Hello fellow humans! There is, and will continue to be, big news around agents, and for most people, it may not be clear what that means, but the AI powerhouse companies are trying to reshape our entire information ecosystem, and the pace of change is breathtaking. OpenAI and Nvidia have both declared that the 2025 is the year of the agent.

Imagine you can just write up a job description and an AI can do it based on your instructions. Anthropic released their new Skills for Claude feature that takes your written instructions and performs them for you. Skills is the hot topic in AI news right now and represents the rapid advancement of AI agents from experimental concepts to practical workplace tools. OpenAI released a similar tool called AgentKit. These developments promise context-specific training and modular capabilities that can handle complex real-world tasks with minimal human intervention. They also mean that you no longer need to able to write code — or even vibe code — to build and agent that performs work for you.

Sources:

We all know how important it is to be able to trust our colleagues, but now AI is a colleague, too. And we have some trust issues with AI right now. Multiple studies and articles are trying to address the critical importance of trust in human-AI interactions. New research examines how trust affects technology adoption, the challenges of "black box" AI systems, and the need for transparency and reliability in AI decision-making processes. This thread emphasizes the psychological and behavioral aspects of human-AI collaboration.

Sources:

Multiple articles are shifting the jobs discussion from AI replacing jobs to AI becoming our colleagues and collaborators. This shift towards AI augmentation focuses on how AI agents are transforming job roles, the need for reskilling, and the balance between automation and human capabilities. How should we understand, identify, and strengthen the skills that remain uniquely human?

Sources:

How do we design AI systems that complement rather than replace human cognitive skills? New research assesses the impact of AI in education. Self-determination theory (SDT) suggests that educators should emphasize autonomy, competence, and relatedness in focusing on critical thinking, problem-solving, and creativity to reduce cognitive offloading. AI tutoring has been moderately successful in STEM and language learning, but concerns remain. Going forward, maintaining human agency in AI-assisted tasks will be critical in the development of intent-based interfaces that understand human behavior and psychology.

Sources:

AI is transforming learning and development processes, both for humans learning to work with AI and for AI systems learning to better serve human needs. This includes adaptive learning systems, continuous improvement mechanisms, and the integration of AI in educational and professional development contexts.

Sources: