Raw LLM Responses

Oh! I can actually provide some perspective here as a previous wet-lab scientist who is in the process of pivoting towards computational bio/techbio/bioinformatics/mlbio whatever you want to call it. Basically Biology is really fucking complicated with a million and one things that can go wrong. Even the most complex protein modelling models (alphafold/boltz/chai/esmfold) have massive limitations and are, at best slightly better homology modellers imo. A lot of "pure" software engineers and mathematicians come in and think they can solve inherently disorder and messy systems, claim the moon and lie (Corso). It's fucking hard. Even leading techbio companies haven't developed therapeutics for novel targets (yet). There's certainty interesting stuff happening at places like recursion or isomorphic but they'll have to validate it in vitro and in vivo before running the gauntlet of clinical trials where 9/10 of all biotechs fail. The field is split into wet lab biologists and chemists who have pivoted over several decades towards ml or mathematicians and software who have done the opposite. Pay is exceedingly variable too, I've known seniors on around 45k at one company but people doing the same job at another can be at 100k+ (UK). Even then that is below market rate for somebody with that background and experience so they'll be given fat stacks to go to places like openai, anthropic, meta, mistral, nvidia etc. After spending around 10 years of your life in education, phds, postdocs, and being paid and treated like dogshit it's hard to say no to that kind of income