Raw LLM Responses

Saying this as someone who works with AI (and not generative AI) in the field of biology, it is an extremely powerful tool for pattern recognition, and it is consistent - minimising the randomness of human error. I've been involved in projects where AI visual recognition has improved our efficiencey by approximately 10x. Now, having said all that, EXISTING professional scientists are prone to chase the shiny bauble that helps them do their work better and faster. The efficiency gains maximise the quality of the output while minimising the cost, improving our chances of winning research grants. On the surface level, this seems awesome. HOWEVER, there is an continuity conundrum that has not been considered well enough. The repetitive synthesis and attention-demanding pattern recognition work is traditionally, what was used to train up NEW scientists, and that kind of mental discipline and attention to detail is what makes truly competent scientists. That kind of work was/is also how many scientists fund post-graduate degrees and fill in employment gaps while seeking more gainful employment. Granted, the truly brilliant 1%ers will always succeed, no matter the environment, but competency is still generally BUILT. That early work shapes a lot of us, and gives us the knowledge and discipline so that we are capable of that creative scientific thinking that AI makes easier for us to do. Dropping a heap of people without the necessary foundational and contextual knowledge into a creative space is unlikely to result in positive outcomes in science or humanity. The 1%ers will benefit for sure, but there will be a net loss of many people who would have BECOME great scientists, but never got the right "soil" to land in. We're also likely to see more of what we see today - a whole HEAP of half-cooked, baseless pseudo-science being pumped out and treated as real science. AI is a great, powerful tool, but that "sword" DEFINITELY needs some careful administration.