Outside In with Charles Trevail

The episode contains a handful of genuine operational data points - AI outperforming coaches on content prediction, 25% annual efficiency gains via LLMs, outcomes parity between chat and video therapy - but large stretches are consumed by personal anecdotes from both parties and a surface-level market overview that adds little for a knowledgeable operator.

“our AI is more predictive than our coaches and therapists are in terms of what content is going to be most engaged with by our members”

“we have increased our efficiency by about 25% a year, north of 5% a quarter, through technologies like this”

A few mildly distinctive framings emerge - the 'organ having trouble' observation and the paradox of simultaneous over- and under-care - but the bulk of the conversation recycles standard healthtech narratives around stigma reduction, step-care models, and prevention-as-brushing-teeth that circulate widely.

“It's weird if you think about it that mental health is probably the only condition where you're asked to use the organ that's having trouble to solve the problem”

“one of the paradoxes of the space is that you have so many people who don't get care, but you also have a lot of people that get far too much care”

Glass is a genuine operator who founded and scaled Ginger, executed a merger-of-equals at ~1,000-person scale, and has direct leadership experience at LinkedIn post-acquisition; his practitioner credentials are real, though the conversational format fails to fully exploit his operational depth.

“I set up a business in 2018 called Ginger”

“when I sold my company to LinkedIn”

A useful cluster of concrete figures appears - 25% annual efficiency gains, named enterprise clients, post-merger headcount, WHO prevalence stats - but claims about clinical outcomes and AI predictive accuracy are asserted without methodology, sample sizes, or comparative benchmarks.

“we have increased our efficiency by about 25% a year, north of 5% a quarter”

“the Starbucks of the world and the Adobe's of the world and the Paramount's of the world, the United Airlines of the world”

The host's prepared T&C challenge is a rare moment of genuine accountability, but it is the exception in a conversation dominated by personal anecdotes, leading questions, and unchallenged assertions about AI capability and clinical efficacy that a sharper interviewer would have pressed on.

“I actually went and had a look at your terms and conditions before this call. So you know where I'm going on this.”

“I love the idea of members first, customers first, et cetera, et cetera”