I recently gave a talk on political risk assessment in strategic decision making for global businesses. While preparing I uncritically ran my notes past an LLM chatbot. It made some comments on the structure, then asked:

Do you want the final part to feel like sober risk management or an alarmist warning about extreme breakdown scenarios?

So now even the bot thinks I'm edging towards the apocalypse. What next. Malcom Tucker style faux praise. Beautiful. Absolutely beautiful.