I was one of the internet’s first influencers. AI just killed the whole category — and created something better

The article argues that AI is shifting online trust away from generic influencers and toward credentialed experts using custom models trained on proprietary knowledge. It cites Dr. Becky Kennedy’s Good Inside platform as an example, with over 100,000 subscriptions and $34 million in gross revenue last year after raising a reported $10 million. The piece is broadly constructive on AI-enabled expert businesses, but it is commentary rather than a market-moving event.

Analysis

This is not an AI demand story so much as a trust-repricing story. The economic value is migrating from generic model access to curated, high-provenance expertise, which should compress the moat of horizontal AI platforms at the margin while expanding it for niche software, vertical SaaS, and creator-led businesses that can package exclusive know-how into subscription products. The second-order effect is that “good enough” AI becomes commoditized faster, but trust-verified distribution becomes more valuable, especially in regulated or high-consequence domains.

The biggest winners are operators with dense proprietary data, brand authority, and a clear accountability loop: healthcare education, financial advice, parenting, legal workflows, and enterprise training. This favors companies that can turn tacit expertise into recurring revenue and human-in-the-loop AI experiences, rather than pure model vendors. It also creates a land-grab for acquisition of practitioner IP, because once a respected expert network is digitized into a domain model, switching costs rise sharply and churn should fall.

The vulnerable cohort is generic influencer media, undifferentiated AI content farms, and ad-supported platforms whose inventory quality degrades as users become more skeptical of provenance. In the near term, engagement metrics may not crack, but monetization could: premium CPMs should bifurcate toward verified sources while low-trust feeds remain traffic-rich but value-poor. Over 6–18 months, the market may start rewarding proof-of-expertise signals the same way it once rewarded follower counts.

The contrarian risk is that the market overestimates how quickly consumers will pay for trust. Most users still default to free, mediocre information until a bad outcome forces change, so adoption of paid expert-AI products may be uneven outside high-stakes categories. Another risk is regulatory backlash if expert-model products imply medical or financial advice without clear disclosures, which could slow commercialization even as demand strengthens.

Key Decisions for Investors

• Long vertical workflow software with embedded expert content, not frontier-model platforms: favor names like DOCS, HIMS, or WIX over pure-play AI infrastructure for a 6-12 month horizon; the thesis is monetization from trust, not model scale.
• Build a basket long of healthcare/financial education businesses with subscription or coaching economics and short an AI content/SEO aggregator basket over the next 3-6 months; expect ad yields and conversion rates to diverge as provenance becomes a pricing variable.
• In private markets, prioritize GP exposure to expert-network rollups and domain-specific AI assistants; target 2-3x MOIC over 3-5 years where proprietary expert IP can be turned into recurring SaaS revenue.
• For public AI winners, use rallies to trim pure horizontal platform exposure and rotate toward application-layer names with proprietary data and liability-managed use cases; risk/reward favors verticalization if provenance becomes a durable buyer criterion.
• Optionality trade: buy medium-dated calls on consumer health or parenting subscription platforms with expert-led brands, funded by shorts in low-trust media/creator monetization names; the upside is a re-rating if trust becomes a paid feature.