No Sloptober

This October, a unique challenge is being proposed: abstain from using LLM (large language model) based tools entirely. It's a personal challenge to rethink our mindset towards these tools. Nuance in language and balance in usage is crucial. The objective is to develop a personal sense of nuance around LLMs, understanding their pros and cons. The idea is to avoid using AI/LLM tools at home or work, relying on the human capability to decrease complexity rather than increasing it.

The challenge also encourages cost risk analysis around LLM usage. It proposes experimenting with reduced or no LLM usage and measuring the impact on team velocity, incident rate, and cost. This could lead to potential savings or risk mitigations in the future. It also suggests applying the "iron triangle" proposition of software development, understanding where one's own gaps are in knowledge and capability, and deciding whether to delegate tasks or automate them.

The challenge acknowledges the usefulness of language translation tools in global communication, but it's seen as an exception where machines can assist humans. However, it emphasizes the importance of professional human translation and localization, which require high accuracy and cannot be delegated to machines.

The term "meat proxy" is introduced, referring to people who use chatbots for work conversations without any thought or minimal editorial review. This is seen as highly disrespectful and a preference for no response. The source concludes by urging individuals to consider what's truly valuable, whether it can be automated, and to adjust suggestions according to personal circumstances while respecting the importance of professional human roles.