Alex Finn discusses the reported return of Claude Fable 5 and proposes using it to review recent code changes, identify possible defects and test a website through a browser. His strong comparisons with other models are personal opinions; his July access and pricing discussion reflects the time of recording.
Alex Finn explores business planning, repeated agent checks of API endpoints and research using public social posts. He also describes building games with development tools. These examples show possible workflows rather than independent proof that generated code or automated security checks are reliable.
For everyday work, Alex Finn recommends listing recurring activities and asking the model which could be automated. He encourages experimentation with prompts and tools; people should inspect the outputs and independently validate important changes and security findings.
Watch on YouTube
