AI scaling is not exponential. It never was. I fit a logistic model to general capability vs compute across every major model generation — AlexNet to GPT-4. R² = 0.98. The asymptote is not infinity. Never was or will be. Thread 🧵
@KRHornberger As a biologist I believe the problem lies with biologists that do not understand Pharmacology. Before joining a drug discovery lab i was told to use the most published compound and not the most selective . @Chemical_probes do a great job, trying to educate on the matter.
Degraders are different!
For years many were understandably skeptical about whether degraders would differentiate from tried-and-true small molecule antagonists.
However, there are now numerous remarkable examples, including from clinical assets, showing that degradation differentiates in many clear ways.
A summary with specific examples here:
Great article. The next REALLY big thing will be using AI and LLMs to generate thousands of new proteins, including new hormones and enzymes and antibodies. @geochurch@demishassabis@_RobToews via @forbes https://t.co/jTOPrpaOKp
What if small molecules could allow you to alter protein localization at will?
What if they could allow you to turn on specific transcription factors?
Our new manuscript reveals bifunctional molecules to do just that. 🧵
💥Our #STI Awareness blog is now LIVE!💥 #STIeducation can be the key to ending the #stigma and more efficiently tackling, testing and treating STIs in the future.
Check out our full #blog here: https://t.co/wvh4JLXj0K
Blog written by Scientist @viviandimou#GIANTmicrobes
Using probes based on the Nobel Prize-winning Click Chemistry technology, we were able to confirm cell-target engagement through subcellular co-localization of our inhibitor and the endogenous HSET protein, on the mitotic spindle.
Congratulations to everyone involved!
Thrilled to share our recent publication from the Centre for Cancer Drug Discovery in The Institute of Cancer Research @ICR_London published today in the Journal of Medicinal Chemistry on inhibitors of the HSET (KIFC1) motor protein. https://t.co/kWeGEXSU0I
Excited to share @VivianDimou 's post for the Mel Greaves award where she has been shortlisted for the third year in a row. As one of my favourite quote says “We are what we repeatedly do. Excellence, then, is not an act, but a habit.”
#cancer#scientificwriting
Unsealing the terrarium🌱
In this blog post, Postdoctoral Training Fellow, @viviandimou provides a unique perspective on the interaction of tumours within the #Cancer#Ecosystem.
This piece was shortlisted for the 2022 Mel Greaves Science Writing Prize.
https://t.co/eipHJJzYCd
Unsealing the terrarium🌱
In this blog post, Postdoctoral Training Fellow, @viviandimou provides a unique perspective on the interaction of tumours within the #Cancer#Ecosystem.
This piece was shortlisted for the 2022 Mel Greaves Science Writing Prize.
https://t.co/eipHJJzYCd
Recently I wrote a post suggesting that peer review doesn't work, and then some weird things happened. A tenured professor threatened to get me fired. Strangers sent me unhinged emails. (People said nice things too.) This week I sort through it all.
https://t.co/nj9svQ5hV5
In my 53 years in the investment world, I’ve seen a number of economic cycles, pendulum swings, manias and panics, bubbles and crashes, but I remember only two real sea changes. I think we may be in the midst of a third one today. https://t.co/aKe9thQnLS
Excited recruitment has opened for multiple faculty positions in #compBio at @ICR_London@CEC_ICR (established group leaders & tenure track). ICR is outstanding environment for basic & translational cancer research; growing #compBio is a priority. Please get in touch to discuss.