I just share an experiment were i use LLMs to predict near future scenarios and a structured "five-wave" framework about technological, economic, and governance transformations, with a timeline...i did this with 8 models, and a meta-analysis
Feel free to play with this
github.com/SwePalm/ai-a...

🤡 On one side, we have the United States government, clutching its pearls about DeepSeek

🕵️ On the other side, OpenAI "Operator Mode"

Who needs state-sponsored surveillance when we have subscription-based screen stalkers?

(reflection by my new Jester style in Claude)

Wrote my first Claude MCP server today...gut feeling is that this concept is great. Will it take of as a standard protocol for connecting LLMs to tool, not sure.
I would like to see adoption of Ollama and LM Studio as MCP Clients, enabling true local execution.
But using Claude on free tier works...

After reaching the "Long chat warning" when using Claude, and I was about to end the conversation, i got this summary, it was a fun journey.