国产成人综合久久免费导航

Dear Keystonians,  

How long has it been since you learned something new about AI? If we aren’t keeping up weekly, we are behind. While that can be overwhelming, I have been simplifying the effort by scanning the landscape for things that have the potential to change my mind about it or upset ideas that are becoming comfortable (that is what learning is, after all!). Last week I heard about a study right in that sweet spot of “Huh! That’s new!”. This study shifts some of the things we thought to be true since the widespread adoption of generative AI just about two years ago this month. (See link here).  

For the last couple of years, the conventional wisdom around AI large language models has been that most white collar industries (knowledge industries like law, finance, medicine, etc…) would not be taken over by AI, but that professionals who use AI would, from now on, be stronger than those who don’t. As many professionals told us, “AI won’t replace lawyers, but lawyers who use AI well will replace lawyers who don’t”.  

? 

A new study from the American medical community recently pushed back on this assumption that we had all been working under. This small but groundbreaking study showed that at the task of diagnosis of complex medical conditions, AI was actually MORE effective than doctors alone, which was not surprising to anyone who has been paying attention. But shockingly, AI alone was also more effective than doctors using AI (and not by a little bit, but by a lot).  

This was the first study that suggested this finding in the medical field and at least one of the authors assumes the finding will be transferable to other fields. The research showed that some of the things that decrease efficacy in human (and human assisted) diagnosis is our tendency towards bias. Doctors tended to outweigh evidence presented by the large language model that supported their hunches and quickly dismiss evidence that was not aligned with their “gut”.  

The researchers found that the cognitive effort of doctors, who study for years to master content related to anatomy, physiology, pharmacology, etc.,?was a barrier for efficacy in diagnosis. In a podcast that I listened to about the study, the hosts asked the researcher what he thought this all meant for people considering the medical field. He hesitated and expressed a deep sense of uncertainty. The author acknowledged that the field was grappling with the implications of this study and the attendant grief that doctors were experiencing associated with the idea of their cognitive capacity not being of much value on its own.  

? 

The researcher also acknowledged that surgeons, technicians, people that do procedures and administer care will (for the foreseeable future anyway) be needed, but he also agreed that the skills that traditional medical school prizes (memorization, for example) are going to keep decreasing in importance in the actual practice of medicine.  

? 

In the end, he said, his top priority is ensuring that his patients get the very best care they can, so he is willing to accept any loss he might feel about the fact that that best care will likely mean huge changes to his job in the years to come.  

This is very exciting for us in education too. We are devoted to ensuring our students get the very best, how will our jobs change to ensure that? We have to pay attention to that question with a great sense of urgency. I am glad to say that at Keystone, we are.  

Interested in the medical field? Make sure you come to our next Education Salon on Monday which will feature an exciting discussion about genetics!  

Hope you all have a great weekend!  

Yours,  

Emily