David Li
@davidycli
Followers
5K
Following
15K
Media
26
Statuses
1K
just a guy trying to make drugs that help patients
San Francisco, CA
Joined July 2014
In biotech, China now ahead in ADCs, bispecific Ab’s, T cell engagers, and traditional small mlc’s US still slightly ahead on cell and gene therapy, gene editing, and more exotic small molecules (glues, covalents, degraders, etc) China clinical dev timelines 50-100% faster
US is firmly, unambiguously losing to China in every area of robotics, every area of transportation other than airliners, every area of energy generation and storage, and every area of infrastructure and construction In 2025 we will lose our lead on AI, semiconductors, and more
48
126
743
Recent advancements in one-shot AI protein / antibody development by Chai, Nabla, AI Proteins, Generate, and a few others are accelerating the *main* theme in biotech: Value of building the molecule is going down. The value of novel targets, novel translational ideas, AND also
**A grand unified theory on what will happen in biotech in the next 10-20 years** the two major forces reshaping industrial biotech in the next decade are: 1. China 2. AI - and they're critically linked how? China's low R&D cost basis democratizes execution by providing
9
22
162
in biotech down markets, what's impt in raising $ is: (having) money >> (team) experience >> asset (differentiation) >> platform reverse in boom times plan accordingly
7
4
59
lots of discourse on here recently on how to get FIH clinical data in US cheaper, faster, and more competitively with China Chinese biotechs, on average across modalities, generate FIH data with ~ $10M USD in funding total (R&D work, clinical ops, mfg, and pt recruitment) but
10
18
136
folks are saying that evolutionaryscale acquisition by CZI has implications for biz model of AI x Bio but I'm not so sure evolutionaryscale was building (to my knowledge) foundation models for sequence / structure / function which are useful for target discovery & translational
7
5
134
Perhaps bullet 7 works over time, but just lol at bullet 6 we are very far from models even beginning to move bottom line on drug discovery (which is clinical probability of success) If they're banking on 6 / 7 transforming the biz model, I think we're looking at trouble
My summary of OpenAI CFO Sarah Friar's vision for the company she laid out today at WSJ tech conference
8
6
63
The implications of this thread for biotech drug dev are enormous. Currently US market generates ~75% of the profit of global biopharma sector. If 1.3bn+ Chinese are truly consuming at same levels of Americans by ~2040, we are about 15 years away from: - China joining US
Well, Americans (and the Japanese and Western Europeans) are still richer on average than the Chinese. But the gap is relatively mild if you compare Chinese living standards to those of very poor countries in Africa and Asia. For example, Chinese life expectancy is 1.3 years
9
17
84
this thread on using ChatGPT to solve an open problem in convex optimization feels like what AI models will act like when they eventually help accelerate science to a "novel" breakthrough - Key part is that human still needs to provide i) pruning feedback ii) prompting on which
I used ChatGPT to solve an open problem in convex optimization. *Part I* (1/N)
0
3
10
✨ Introducing Benchling AI — the first command center for scientific #AI. Built-for-science agents, models, and your R&D data, all in one trusted place, so it’s right in your daily workflow. #BenchlingAI is available today for all customers and free for academics!
0
0
1
AI for science space continues to heat up - but are we approaching the end of the beginning? ie have the breakout AI vertical leaders for science already been anointed like in coding, customer success, legal, etc? I think not. But, we are getting closer - First a debrief of
8
2
49
lynbrook, mission sj, lowell --> silicon valley pipeline is west coast equiv of stuy, bronx science, bergen county --> wall st pipeline same status games, different coast
as someone who went to both, can confirm 🤠 lynbrook was a public high school where being smart was cool and there was no substitute to working hard. the students were extremely disciplined and took nothing for granted it's probably not a coincidence that the founders of
1
0
4
Silicon Valley “gangster story” lore: Steve Jobs spends $50M over 10 years to buy and personally fund Pixar, creating an iconic animation company that sold to Disney for $7b Biotech version: Bob Duggan spends $50M out of pocket to buy and fund Pharmacyclics over 8 years,
Replit CEO Amjad Masad on the “most gangster story in Silicon Valley” In September 2021, Replit founder Amjad Masad tweeted: “The most gangster story in Silicon Valley is Steve Jobs buying Pixar for $5m, investing $50m, operating at a loss for a decade — so much so he had to
0
2
26
Starting a software co is like building a plane & trying to get it to lift off before you run out of runway Starting a biotech co is like trying to solve an escape room before time runs out For software co, getting the plane off the ground requires primarily ppl who can work
Keith Rabois: “I tell founders not to worry about runway. Worry about lift.” “If you think about lift in a plane context, a company is only valuable if you achieve lift. Runway is a tactic for achieving lift, and you may need to extend the runway so that you have more time to
1
8
63
growing crop of biotech newco's that are leveraging china preclin execution (not just CROs) coupled with US translational bio + clinical trial design mostly working in genetic medicines, but also in ADCs and antibodies funded by mostly non-US biotech VCs, although a couple
9
9
98
Only two types of founders getting funded right now in biotech 1. phd drop-out with touch of tism working on longevity 2. multi time biotech CEO who has returned $ bn's to specialist biotech funds sorry, don't make the rules around here
26
10
240
just back from shanghai / beijing / HK, met several chinese pharma, 10+ biotechs, local VCs - notes / comments from the trip: * what China did to ab's / bispecifics / TCEs, they are about to do (or already have) to genetic medicines * there are now 120+ "in vivo CAR-T"
12
47
384
The generalists are back 🙌
We should be investing far more into AI-driven drug discovery. Every extra unit of compute has the potential to translate directly into more lives saved (and can have exponential returns). That’s the purest expression of what technology is meant to do.
0
0
15
the existence of REGN sets a high bar for AI bio tx startups Looking at 6 (and likely two more) approvals out of 13 early clin assets from 2017 pipeline proof positive that genetics focus + high level drug discovery execution can vastly outcompete industry avg (~20% successful
In light of recent phase III successes (REGN1908-1909 and garetosmab), it's kind of interesting to reflect on the $REGN pipeline of 8 years ago when these two were in phase I. Of 13 early/mid programs: 6 (46%) have been approved in their intended indications: Dupi,
3
1
53
lots of folks comparing AI revolution to microprocessor wave in 70's people tinkered in garages, many PC startups became trillion $ co's @ganeumann makes the opposite argument here - what if AI is more like container wave of 60's - lots of capex and incumbents "won" rather
1
3
11
