Amit

Building a new operating system for research funding

[Amit @AmitRnD]

Science is humanity’s highest-leverage investment.

[Amit @AmitRnD]

It's impossible to use Fable 5 every time I want to do anything I get dropped down to Opus. Extremely frustrating.

[Laura Minquini @LauraMinquini]

I don’t know who needs to know but the problem in drug development or therapeutics is not lack of hypothesis.

It’s proving them.

If you don’t have the infrastructure, the right data or development capabilities, no amount of AI, compute, or agents it’s going to solve this.

[Amit @AmitRnD]

@NikoMcCarty Super cool

[Amit @AmitRnD]

@aaronmring Not using AI at all is not an option. Those that go down that path are doomed

[Amit @AmitRnD]

@aaronmring They are going to get into every business that's just a function of dollars and neurons. Everyone needs to assume that the frontier labs are going to train on your data. The question is 1) does that really affect your business and 2) if so, how are you going to self-host AI?

[Amit @AmitRnD]

I except anthropic to go into drug development. It's a very tractable problem and they have the balance sheet to stomach it. The only problem is it's generally a lot less profitable and less certain than SaaS, but why else buy @CoefficientBio

[Amit @AmitRnD]

IMO Anthropic will be successful, as a biotech company, but won't transform medicine or biopharma.

The first thing to realize is: Most biotechs aren't discovering, manufacturing, studying, or selling drugs

[Adam Feuerstein ✡️ @adamfeuerstein]

AI company Anthropic announces it will begin developing drugs of its own.

Preclinical work in rare diseases.

https://t.co/xTmpDK79uI via @brittanytrang

[Amit @AmitRnD]

The things biotechs do is provide judgement, expertise and coordination - AI excels at. The other piece of the pie is raising lots of money and frontier labs are much better at that than biotech and have seemingly endless supplies of low-cost capital.

[Amit @AmitRnD]

Hallelujah

[Anthropic @AnthropicAI]

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.

We'll begin restoring access tomorrow, and will share an update soon.

We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

[Amit @AmitRnD]

Any investors interested in backing a Sanhujoriwon in Guam asking for a friend...

[Amit @AmitRnD]

A lot of rare diseases are pretty trivial to cure. We know how to alter gene sequences (CRISPR, Base Editing, Prime editing) and we also know delivery - LNP isn't perfect but there's tons of other tissue-specific NP platforms. The real question is are we willing to try them out!

[Ron Alfa @Ronalfa]

This is great. So many rare diseases that could use some impact. I think this can actually be scaled today.

[Amit @AmitRnD]

The problem is never finding drugs or targets - it's deciding which one is better than all the rest
(before you're already in Phase 3)

[Ron Alfa @Ronalfa]

"It all comes down to the hard problem of is this the right target for this disease." @VasNarasimhan

[Amit @AmitRnD]

Extremely exciting launch (watching now) also another point towards why just wrapping Claude (or ChatGPT) in prompts, harnesses, and MCPs is a losing bet for big SaaS markets.

[Adam Feuerstein ✡️ @adamfeuerstein]

Anthropic releases Claude Science, a product aimed at researchers, the pharma industry https://t.co/sGPS7Aj2gI via @matthewherper @brittanytrang

[Amit @AmitRnD]

Shadow trials are an extremely interesting method to get real clinical data from people - creating a public good by self-experimenting with new drugs and sharing their data. It's also an amazing case of creating economic alignment and using prediction markets for useful things

[Michael @endpointarena]

Endpoint Arena is building a mobile app called Anecdote Bio.

The problem with AI is that it rehashes the same public information to predict trial outcomes.

The real alpha is data in your phone. Here is how it will work and how you can make money. 🧵

[Amit @AmitRnD]

@johnarnold Not just $15m; greater of $15m or 10x investment and that goes to every downstream LP. It's possible the returns for the entirety of the first $75m in investment are all tax free