#06 – D.A. Wallach: music, medicine, longevity, and disruptive technologies

KEYPOINTS

• 01

  Explore an instrument freely first, then learn theory.

  Wallach argues that theory should answer questions you have already run into while messing around on an instrument, so it never 'imprisons' you. His plan for teaching his future kids piano is a daily 30-minute sit at the piano with no rules. Anyone sitting there that long will start playing, and after tens of hours of exploration, theory feels like a gift because it explains what you've already found.

• 02

  Singing is musical speaking.

  Wallach had never sung before college. He tried out for Chester French as the drummer, lost the spot to his friend Damien Chazelle, who later made Whiplash and La La Land, and became the singer as a 'consolation prize'. Once touring demanded an hour or two of singing every day, vocal coaches taught him that speech already has pitch and singing is just controlling it deliberately, which made singing much easier for him.

• 03

  Chester French broke through with recordings, not live shows.

  The usual path of playing campus parties and then touring East Coast colleges failed for the band. After sophomore year they decided their real product was recorded music and spent about two and a half years in the studio making one complete album. They sent it out midway through senior year, it reached Kanye West, and they eventually signed with Pharrell and Jimmy Iovine's Interscope. Wallach thinks they were a musician's band whose fans were mostly people who played music themselves, and that the internet was just then disrupting how labels found artists.

• 04

  Wallach is skeptical of elite universities.

  He studied African American Studies with a minor in Kikuyu, and calls Harvard essentially a 'vocational school' for investment banking and management consulting. He estimates 40–50% of his classmates went that way, and partly rebelled by committing to music. He argues these schools reproduce social hierarchy and turn unique high schoolers into 'boring high-achieving technocrats'.

• 05

  Spotify succeeded because the music industry had already been wrecked.

  Wallach says piracy through Napster and CD burning shrank the record business to about a quarter of its former size. Sean Parker and Shawn Fanning had wanted to legalize Napster, but labels were too defensive. It took decades of decline before rights holders were desperate enough to license something like Spotify. To Wallach as a customer, $10 a month felt 'as if I bought everything ever', unlike Pandora's algorithmic radio, where you couldn't pick songs.

• 06

  Healthcare has barely been disrupted, and Wallach sees no obvious 'Spotify of healthcare'.

  Most healthcare money goes to labor. Hospitals are not great businesses, and he cites Kaiser's roughly 3–4% margin. Insurers have legally capped margins. Drug companies earn large profits, but they are the ones taking the risk to invent new treatments. He expects many small efficiency gains from consultants, a shift toward technology plus highly skilled nurses and fewer scarce physicians, and the biggest gains from new medicines.

• 07

  New medicines are what give people more years of life.

  Wallach turned from fixing the US care-delivery system, which is important but local, to biotech, because 'the biggest game changers in human health in the past 50 years have been pills.' He describes precision medicine as treating disease at the level of mechanism with little collateral damage, and says we are in 'inning one or two'. He says his personal motivation is reducing existing suffering more than extending lifespan.

• 08

  Wallach's 'singularity' is the ability to simulate biology digitally.

  Today biology is learned one variable at a time in mice, dogs, monkeys and people. He calls clinical trials 'a brutal form of learning' whose participants are heroes. Systems biology, which turns mechanistic diagrams into formal code, is a waystation toward simulating biology and running experiments at zero marginal cost. He doesn't know whether that is 20 or 1,000 years away.

• 09

  Finding tumor DNA in blood is like finding 'a needle of hay in a haystack'.

  Prenatal blood screening works because roughly 10% of the DNA in a pregnant woman's blood can come from the fetus, and half of it is foreign, coming from the father. Early tumor DNA may be under 0.5% or even 0.1% of the total, and it is the patient's own DNA with only a few mutations. Grail, spun out of Illumina, tries to find it by sequencing very deeply against known cancer genes.

• 10

  Attia suggests stacking screening tests like layers of Swiss cheese.

  Each test has holes. Mammography sees calcified lesions, and diffusion-weighted MRI sees most of the rest. Attia would layer those with a liquid biopsy and a Bayesian layer of family history and mutations, so few cancers can slip through every layer. He thinks liquid biopsies should confirm findings rather than lead. Wallach pushes back that leading with them depends on their sensitivity and specificity, and both agree no test is perfect on both.