I've been wondering whatever happened to this since they went to the Anderson cancer center. Working in health care in the data field, there's a lot of low hanging fruit that humans can do very...
I've been wondering whatever happened to this since they went to the Anderson cancer center. Working in health care in the data field, there's a lot of low hanging fruit that humans can do very effectively very quickly. However, some of the predictive space can be automated (the example of calling all 45+ yr old men for a prostate exam is a good example) for population health benefit. It's the insurance providers not the place of care, however, that will be interested in something like that.
It seemed to me the primary value of Watson would be in helping human doctors stay up to date on current discoveries, research, and treatment. A black box where you submit patient records and get...
It seemed to me the primary value of Watson would be in helping human doctors stay up to date on current discoveries, research, and treatment. A black box where you submit patient records and get back a google-style search result of what it might be. What the human doctors think will be on that list, but there might be some other information they aren't aware of that Watson learned yesterday and can now share with them. Watson should be amplifying doctors' knowledge and saving them time, not doing their jobs for them.
There's a general rule that I have with automation technology of any kind: there should always be a human to review the results of the automated work. That is, you should automate where you can,...
There's a general rule that I have with automation technology of any kind: there should always be a human to review the results of the automated work. That is, you should automate where you can, but you should always manually validate the work being done for defects.
In this case, blindly trusting a list of results would be stupid, but reviewing the list to ensure that everything you expected to be there is there and then reviewing the possibilities given to you would be of great benefit.
I've been wondering whatever happened to this since they went to the Anderson cancer center. Working in health care in the data field, there's a lot of low hanging fruit that humans can do very effectively very quickly. However, some of the predictive space can be automated (the example of calling all 45+ yr old men for a prostate exam is a good example) for population health benefit. It's the insurance providers not the place of care, however, that will be interested in something like that.
It seemed to me the primary value of Watson would be in helping human doctors stay up to date on current discoveries, research, and treatment. A black box where you submit patient records and get back a google-style search result of what it might be. What the human doctors think will be on that list, but there might be some other information they aren't aware of that Watson learned yesterday and can now share with them. Watson should be amplifying doctors' knowledge and saving them time, not doing their jobs for them.
There's a general rule that I have with automation technology of any kind: there should always be a human to review the results of the automated work. That is, you should automate where you can, but you should always manually validate the work being done for defects.
In this case, blindly trusting a list of results would be stupid, but reviewing the list to ensure that everything you expected to be there is there and then reviewing the possibilities given to you would be of great benefit.