Reading all these stories about rogue #AI agents doing bizarre things on their own, which the users only find out after the fact, makes me wonder why that's even possible.By which I mean, if it was a fresh junior developer I would have them give me a plan of what they are to do and I'd expect them to work that plan or tell me if it needs to be changed. Moreover? If anything in the plan looks hinky I'd want them to check with me before starting that step.Why are AI agents given total freedom?
Related
Xiaomi's new #opensource, #agentic #AI coding harness MiMo Code beats Claude Code at ultra-long, 200+ step tasks - https...
Xiaomi's new #opensource, #agentic #AI coding harness MiMo Code beats Claude Code at ultra-long, 200+ step tasks - https://venturebeat.com/technology/xiaomis-new-open-source-agenti...
European Commission publishes Code of Practice on marking and labelling #AI-generated contenthttps://digital-strategy.ec...
European Commission publishes Code of Practice on marking and labelling #AI-generated contenthttps://digital-strategy.ec.europa.eu/en/news/commission-publishes-code-practice-markin...
🐧 15 Best Free and Open Source Status Page SystemsA status page system is software that lets you communicate incidents, ...
🐧 15 Best Free and Open Source Status Page SystemsA status page system is software that lets you communicate incidents, schedule maintenances and downtimes with your customers. The...