Daniel Brox’s Post

Wow. Maybe the next step, picture + voice sample + text script = realistic video? Getting the voice script itself is a lot of work!

Fascinating summary of how the ACNU pathway may change part of the pharma landscape, hopefully for the better for all…

Be safe y'all. I received a message from a person who indicated his boss at Warburg Pincus LLC was interested in networking with me. WP makes large investments in life sciences, so I was initially intrigued (appeal to ego?). He asked for my phone number so his boss could follow up directly. Several things seemed fishy. When I asked a few questions trying to verify the person was legit, he deleted his account. Would be happy to share what the clues were, but don't want to help people improve their scams. #onlinenetworking #security #trustbutverify #businessethics

Looking forward to the HBS case which will one day be written about this. In the meantime, grabbing my popcorn to see it play out over time.

Great article for my clinical stage biotech friends.

I have recently been thinking about the broader implications of AI on society. I think AI may begin to unwind the core of post modernity, the notion objective truth doesn't exist. As with any model, the better the data, the more useful the AI. Data scientists are going to spend decades codifying domain specific ontologies based on what we know to be true to make the models produce better results. No one wants a model that hallucinates. Trustworthy AI is truthful AI. Truth will become a shared project again. Whether you agree, disagree, or think I'm hallucinating, would love to know what you think.

Happy 40th birthday ZS! I am extremely grateful to have learned about ZS in 1999 from James Yen which was the beginning of what has been a fantastic voyage. Along the way, I have made so many dear friends. Many of these friends have remained at ZS, but many have gone on to do other exciting things. When I think about ZSers, I think about the 42K current colleagues and alumni who have each made contributions to our clients, our people, our firm, and to our network of friendships. #ItsStillThePeople!

I found this report by colleagues Pratap Khedkar, Daniel Reiss, Alex Anokhin, Jessica Jarvis, Kapil Pant, Arun Shastri, and Nicolas Zimmerman, along with multiple collaborators at WEF to be both inspiring & practical. In this report, you can find how experts scored 18 use cases based on their potential for global health impact and need for public-private collaboration along with deep dives on several of the most interesting use cases, barriers, and ways to overcome them. There is a bit of something for everyone in here. But for me, I really was inspired by: 1) The power of data driven diagnosis and risk stratification for patients. This concept was the thing that turned me on to RWE 15 years ago and it still excites me today. 2) The use of federated learning models to overcome challenges with data sharing/privacy. But that is just me. What inspired you? https://lnkd.in/eibjnWU5

I started playing around a bit with ChatGPT when it first made a big splash in the media and I was immediately amazed at what it could do at the time (e.g. write a clinical trial protocol for an Alzheimer's drug, write a winsome letter to a client to address an issue). I was also shocked at what it could not do (e.g. write a haiku about the joys of running). When I shared this with a friend (Paul Maloney), his response was "give it a month". Well, I gave it a month and now it can. One of the fun things about ChatGPT is that you can ask it to "do it again" and "again". Given how modular haikus are, if I combine its three attempts, I arrive at: Beneath morning sun Liberating wind's embrace Bliss found in each step (I was struck by the apostrophe. I think it works with and without, but with personifies the wind in a more creative way, no? Way to go ChatGPT. :-) On a more professional note, I have also used it quite a bit to bone up quickly on a variety of disease states, to quickly understand the standard of care, burden of illness, unmet needs in the market, etc. I do take all results with a pinch of salt. Always good to apply Reagan's advice to "trust but verify". But I would say most of it is holding up really well, and definitely saved me alot of time.