The future of medicine isn't a single technology - it's a collision. AI is automating the clinic, while older interventions show alarming long-term effects.
On This Week in AI, Shiv Rao argued that large language models will soon handle routine consultations for conditions like rashes and colds. The real transformation, however, is in administrative burden. Doctors need 30 hours a day to complete required tasks; AI agents will coordinate the entire care continuum, from intake to post-visit orders. Rao’s answer was clear: consult top AI models before seeing a lower-tier GP.
This efficiency push contrasts starkly with findings on TFTC: A Bitcoin Podcast. Nicolas Hulscher presented data showing mRNA vaccine components, including cancer-promoting DNA, can persist in patient tissues for over three years. He cited population studies linking the shots to increased risks for seven major cancers and over 136,000 excess U.S. deaths since 2021.
Meanwhile, on the cutting edge, This Week in Startups reported that Paradromics is implanting its first human patients within weeks. The goal is to restore speech by reading neural activity, with AI cleaning up the signal. The same principle could extend to motor control or reconstructing sensory experiences.
Beyond hardware and pharmaceuticals, Huberman Lab highlighted a simple, proven mental tool. Dr. Richie Davidson noted randomized trials show just five minutes of daily meditation for 30 days reduces depression, anxiety, stress, and inflammatory markers like IL-6. The practice builds resilience by observing stress, not avoiding it.
The Peter Attia Drive added a crucial philosophical layer: happiness is a durable state built through practice, not a fleeting feeling to chase. The key is using metacognition to control aversive experiences, turning chosen suffering into strength.
Healthcare is splitting into parallel tracks: one deploying AI to solve systemic inefficiencies, another confronting the unintended consequences of past technological leaps, and a third exploring the profound potential of directly interfacing with the brain.
Shiv Rao, This Week in AI:
- I would always do the models and then figure out who to see.
- Two choices: go to the lower third of a general practitioner's and get advice, or get it from the top 3 or 4 models.




