social informants solutions

Lyd and Albert are brainstorming on the use of the OpenAI model to train ai to play video games. AI can operate on cases that are too complex for human response time and understanding. Share the examples

OpenAI did research on training AI to play Minecraft and made observations as to how well the system performed in training and how quickly it was able to achieve new items. As Lyd was reading it, she started making immediate comparisons to healthcare and could see ways that the technology could be used in surgical settings.

In 2018 NIH predicted the following impacts by AI, there is a shift in language but conceptually the NIH researchers were able to accurately think through the likely opportunities from 2021-2024.

This will improve access to healthcare in a number of ways. Complex cases will be seen. More people will be seen by providers while the ai does work in operating rooms.

people with complex tumors that sit in the brain or are wrapped into blood vessels. infants. pregnant women.

ease on the workforce issues: not enough doctors to go around in primary and specialty care. people live longer. being a doctor has gone out of fashion.

AI is expensive at scale. Lyd needs help explaining this, the math in the articles is confusing because we use tokens and some amount of data take a small part of a token but that is converted to cost. Normal people cannot understand it. I cannot understand it and I am not normal. Descriptive terms and allegorical descriptions are needed.