I've been finding that using the various LLMs out there, and building RAG solutions, that it is a matter of time until there are tools that enable an average person to consume much more information, work towards new research strategies, and over time LLMs will understand the science at a much more fundamental level, it is already capable of simple logic, much more complex logic and probabilistic reasoning is possible.
Integration with other AI tooling can make this especially powerful. Like alphafold.
LLM based AI is already proving very good at summarizing vast amounts of text. Combined with scans of latest published research, one can stay on top of the research, with incredible summaries and ranking of what is important to look over.
Understanding of the biotech world (LLM base models, and continued fine tune pretraining), can make it easy to create an AI experience that guides someone through the pipeline of getting research into production. Help with funding, managing a team, how to run a biotech startup.
My Thesis Here Is: in a matter of a few short years, AI experiences will be available to an average person without any funding, can use an AI that can guide research, help manage connections to industry players, how to run a bio-tech startup, how to profit.
Anyone want to build this? I'm willing to give it a shot.
[link] [comments]