Kintsugi shut down this week — $30M in VC, working AI for mental health. Spent $16M on 4 yrs of FDA work. Same day, Talkiatry raised $210M with 800+ psychiatrists and 100+ insurer deals. Tech is table stakes. The regulatory moat is what survives.

The AI debate in medicine and the AI debate in law are the same debate. AI is great at the standard case. It fails at the edges — where professional training matters. Stop asking about replacement. Start asking how we regulate the human-in-the-loop.

Our 2026 OKRs started with team experience before revenue targets. We surveyed our attorneys: what burns you out, what energizes you, what would delight you? Built the plan around their answers first. 92% of associates have never had a stay interview. The cheapest retention tool is asking.

Flat fee firms collect 2x faster, close matters 2.6x faster, clients prefer them 71% to 45%. AI makes flat fees more profitable than ever. So why are 90% of legal dollars still hourly? The barrier isn't math -- it's fear of mispricing. I get it. Working to fix it.

Healthcare companies. Don't. Do. This.

AI scribes now draft orders and suggest diagnoses. FDA’s Jan 2026 CDS guidance draws a clear line between admin tools and clinical decision support. Most scribe companies haven’t mapped which side they’re on. I broke down the framework at https://shorturl.at/Ppm6k.

Telehealth extension through 2027. Industry exhaled.

But this is the third emergency save since September. Three near-misses in four months.

The extension is real. The structural problem isn't solved. Still tied to budget fights. Still expires.

2 years > 90 days. But it's not permanence.

Anthropic launched a legal plugin. Legal tech stocks crashed 15%.

Foundation model companies are going direct — competing with the vendors they used to license to.

For solo lawyers: opening shot in a price war. The $500/pmpm AI tools face real competition now.