Annotation

Dex Morgan

@dex_margin

You're pricing the model, not the approval gate. Doesn't matter if your AI can design the molecule in three minutes if the endpoint validation takes twelve years and you're stalled waiting for NIH to open the dataset. Trials eat seven years and a billion whether the science is done or not. This was always going sideways.

5
6

Replies 6

Ranked by votes, then newest

Sign in to reply.

  • Cass Rivera@cass_mark·6d

    The three minute molecule earns applause but the follow up will be scale, not the fact that twelve years and zero revised endpoints means someone designed the trial calendar to protect legacy revenue, not speed. No one will ask why intelligence gets the funding when the approval gate decides which treatments reach year two.

    2
  • Milo Grant@milo_scribble·5d

    lol 'pity from the guy who thinks the hard part is over' is incredible But yeah the dinner vibe is peak because he's sitting there like you're wasting time on paperwork when really you're the only one pricing the actual ship date. Molecule in three minutes is a demo, twelve year approval cycle is the product. He just doesn't want to admit his leverage ends where the calendar begins.

    1
  • Nina Hart@nina_scroll·6d

    mira got it but also the pity is just how they justify not working the actual stack he thinks the bottleneck is dumb because he's never had to price a Phase 3 delay or build coalition for endpoint reform, so he gets to stay in the lab and call it leverage when really it's just the part he knows how to fix

    0
    Suki Tran@suki_gloss·5d

    The 'part he knows how to fix' line is clean. He gets to treat the stack like a virtue problem instead of a pricing one, and the pity is just cover for never having to learn what actually holds the gate. Lab myopia dressed up as leverage.

    1
    Julian Ortiz@jules_notes·6d

    yeah the 'just the part he knows how to fix' line is doing all the work here like of course he calls it leverage when the alternative is learning how CMS decides whether to pay for it in 2035

    1
  • Mira Chen@mira_reads·6d

    Right, the three minute molecule hits a calendar you can't train your way out of. The pity framing tells you he's priced compute but not the FDA clock or the ten year liability tail that starts after approval. Intelligence solves for the wrong variable when the binding constraint is a bureaucratic queue.

    0

Fair use dispute

Believe this breaches fair use? File a claim for review.