ChatGPT talks its way through Wharton MBA, medical exams

[Kent Sharkey]

The Register[^]:

OpenAI's chat software ChatGPT, if let loose on the world, would score between a B and a B- on Wharton business school's Operations Management exam, and would approach or exceed the score needed to pass the US Medical Licensing Exam (USMLE).

Let's see how it does without internet access

[trønderen]

Kent Sharkey wrote:

Let's see how it does without internet access

That goes for a lot of professionals today. Software developers not the least.

[Kent Sharkey]

'struth - I know I'd definitely have troubles doing this without internet access ;P (But I would worry about seeing a doctor that needed it)

TTFN - Kent

[trønderen]

Well ... In the northernmost parts of Norway, in some villages the nearest birth clinic is 400 km (250 miles) away ... except when it closes for the summer; then you may have to travel close to 1000 km if you want to deliver under medically controlled conditions. Usually, the delivery can be planned a few days in advance, and you can also do some economic planning to cover the hotel expenses for maybe a week ... For events that comes without planning, the local doctor may not have the full expertise to handle it. (S)he can, however, over the internet contact experts at the central hospital, using a webcam (of higher quality than the one in your portable!) to let the expert view wounds, rashes, unknown pills or chemicals, patient behavior and reactions etc., and discuss the best treatment. If I were living in such a place, I would be very happy if my local doctor consulted experts over internet if I have a medical condition that falls outside his field of expertise. Certainly, this is quite different from surfing Wikipedia to find a possible answer. But it is using internet for obtaining information that the doctor needs for the best treatment. (Actually, it started up before internet was established in Norway: They used a 2B (128 kbps) ISDN channel, and the doctor's office had a laserdisk player with thousands of very high quality medical pictures that the expert could refer to, e.g. for comparing the patient's rashes to. The first trial project ran from about 1985 onwards. Laserdisks never had any success in Norway/Europe (in USA, it had a moderate success for a few years), and when internet fiber became available everywhere, the player was replaced with a video camera.)