Beyond High Performance

A handful of genuinely useful framing devices emerge—training AI how to treat you, the cast-versus-weights distinction, and the regression-to-mean observation on AI wisdom—but they are surrounded by extensive philosophical meandering, film references, and soft motivational language that dilutes the signal-to-noise ratio considerably.

“You are always training whatever AI you're using how to treat you, whether you realize it or not.”

“LLMs are trained on all data, which means it's going to be a return to the mean idea of wisdom. Actually, I think most people, including myself, are less wise than the gradient mean.”

The cast-versus-weights and CrossFit-versus-Ozempic analogies are fresh and useful, and the 'service slop' framing is ahead of common discourse, but much of the AI-and-humanity conversation recycles well-worn McLuhan, Jurassic Park, and WALL-E references without adding a genuinely novel argument.

“Every increase in technology demands an increase of character to meet it.”

“Service slop. To where you can, like get customer service. But you, you know, you're not talking to a real person now.”

The participants are internal coaching-firm partners at Novus Global—practitioners with real client experience—but this is an in-house roundtable, not an external operator who has deployed AI at enterprise scale; the Sidekick product is still in development and the discussion stays at the level of philosophy and anecdote rather than demonstrated at-scale execution.

“Tricia has been running a skunk works for the last year or so around... Sidekick”

“I will coach some of my GPTs to treat me a certain way.”

The episode is almost entirely abstract and anecdotal; the most concrete moment is the story of a meeting recording sent without a 'kindness filter,' but there are no named client case studies, no metrics, no adoption numbers, and no verifiable data points about the Sidekick product or its outcomes.

“So we had a pretty frank conversation... Instead what he did was he sent a recording of the video to them to watch.”

“That tool joins and listens, and it listens with an awareness of all of the past sessions.”

The host surfaces a few productive prompts—pushing on how you'd know if you're using AI as a cast versus a weight, and drawing out the Sidekick story—but the format is a collegial internal roundtable with consistent mutual validation and no genuine pushback or challenged claims throughout.

“That's really good, Tricia.”

“How would you know if you're using it as a cast or as await”