Kenkoko said:
irvinehomeowner said:
?Every? clinic may not want to share data nor can afford a 7-figure price.
Here's the reality with your "every" clinic take.
Medical Private practices are dying very quickly. When I got my license in 2008, 65% of doctors were in private practice. That number has dropped to below 25% today. And the trend is continuing because we are heading to single payer system or a robust public option. Our policy makers believe large clinics are more data-driven, more efficient, and more effective than their small, private-practice counterparts. So laws and regulation are being pushed this way namely the Affordable Care Act. Very soon small private practices will basically go extinct and frankly 7 figure is a drop in the bucket for any med to large clinic.
Sure, but my take was because IC used the word "every" which in reality is not the case due to cost, lack of connectivity, etc etc. You are referring mainly to US but what about the rest of the world?
You also missed my first point. Software companies make it very inexpensive to get clinics started. (some are even free!)
Their goal right now is to get massive adoption not immediate profit.
I did not because I think you missed mine. The software is not the only cost. There are many other components required for this to work in a widespread manner. While you may have experience with AI usage, I have experience with infrastructure development and I know enough about AI that it's not like you just say "Hey Siri" and now every clinic can get an in-depth image analysis from an MRI.
In my opinion, it is bad long term business practice to share data. But it is hard to fight progress and profit. Therefore, the overwhelming majorities do it. Because they are in the business to make money. Every dollar saved = dollar in pocket.
I think this depends. Medical data is especially subject to privacy laws so there has to be caution here.
I think my boss is slowly awakening to the realization that we are becoming too reliant on the software. The next time we need to upgrade, it's going to be painful.
This is true for everyone. Most of us can't go through the day without Internet (TI being an example
).
But at the same time, with progress there are pain points. For example, everyone wants cloud applications, however, that type of environment is inherently less secure that an application on your internal network. People want more access but that also gives more access to hackers. I'm sure everyone of us knows someone who may have got ransom hacked. The more accessibility we have to data, the less privacy can be guaranteed.
Which also makes me wonder... if AI is just software... it's subject to being hacked. How do we know in this example of radiology, that the diagnosis will be correct or if the AI has been compromised? I like to kid about Terminator scenarios but what security is being employed such that AI does not harm more than help?
So as much as all those forward thinkers worry about replacing jobs and affecting the economy, a just as important concern is making sure that AI doesn't go the wrong way (which I can ensure people will try to do because bad people exist). Every tool can become a weapon.