AI Agents Are Here. This Is What They Can Do—and How They Can Go Wrong
1 min read
Summary
AI ‘agents’ are the next iteration of the large language models that run ChatGPT and Copilot, with more sophisticated functionalities, enabling teams of AI models to work together.
Developed by OpenAI, Projectrend, DeepResearch, and ChatGPT are all “agents”, with the latter described as the first step towards an ‘AGI’ system that can think and act.
Uses for agents include working together to carry out complex research tasks, or being deployed by a company to carry out multiple tasks such as writing meeting notes and handling defect management.
The main risks are human displacement from jobs and over-reliance on AGI, which can lead to errors and inaccuracies that result in harm.
