The people creating AI are often hyper-intelligent and sensitive, yet they may not be prepared for the intense scrutiny and expectations placed upon them.

We tend to give up on AI models much faster than we would on human employees, despite the need for patience in both cases.

The expectation versus reality of AI adoption is often skewed; while expectations are high, the actual capabilities of current models may not meet those expectations, leading to a rapid adjustment in perception as models improve.

Outperformers treat AI like a teammate, while underperformers treat it like a tool, which fundamentally changes the outcomes they achieve.

While AI is incredibly useful, it still requires human oversight for critical tasks due to its tendency to make errors.

Trust in AI systems is becoming more critical than raw performance, as users need to feel confident in the reliability and consistency of these systems.

AI is trained on averageness, while each individual possesses a unique sensibility; ceding our preferences to algorithms risks losing our agency.

AI is increasingly capable of handling transactional interactions, often exhibiting more patience and nuance than humans, which challenges previous assumptions about AI's limitations.

There's a huge premium to naivety in AI; approaching industries with a blank slate can lead to accomplishments that others find difficult due to their preconceived notions.