~/ARTIFICIAL I/ethan-mollick-s-ai-guide-the-shift-to-agentic-systems

Ethan Mollick's AI Guide: The Shift to Agentic Systems

Simon Willison highlighted Ethan Mollick's updated guide on AI tools, which marks a significant transition from simple chat-based LLMs to agentic systems capable of executing complex, multi-hour tasks. The guide notes that Google's Gemini has fallen off the list due to the lack of an established competitor in the computer-access agent category. This shift reflects the broader industry trend where AI is evolving from passive assistants into active agents that can control computers and run workflows autonomously. Understanding these new agentic modes is crucial for users looking to leverage AI for high-productivity, end-to-end tasks. The guide highlights confusing naming conventions, such as ChatGPT's "Work" and "Codex" modes versus Claude's "Cowork" and "Code" modes, which function differently depending on whether they are used on mobile or desktop. Notably, switching ChatGPT mobile to "Work" mode allows its Code Interpreter to access the internet.

## BACKGROUND

Agentic AI systems are designed to pursue goals through autonomous actions, such as browsing the web, calling APIs, or editing files, rather than just generating text responses. Tools like OpenAI's Code Interpreter and desktop integrations allow these models to execute code and interact directly with a user's operating system to perform complex workflows.

## REFERENCES

## KEYWORDS

#Artificial Intelligence#AI Agents#LLMs#Productivity Tools

$ subscribe --daily

Ethan Mollick's AI Guide: The Shift to Agentic Systems | Daily News