Is anyone concerned about the influence of artificial intelligence on their career as a pathologist?

[Asteroid Body]

I'm not. I'm also not able to disclose exactly why, or who I work for, but I can say that I've developed an appreciation for the unique contextual visual recognition skills I've developed as a mediocre pathologist, while training deep learning algorithms as an undercompensated side hustle.

I've never been able to beat the computer at chess. However, very good chess players still beat computers.

Pathology residents seem to learn more quickly than deep learning algorithms.

---

Members don't see this ad.

 

[Unty]

No.

 

[B***F*****]

Not worried. I think by the time it might have an impact, I will have retired. This will impact a future generation of pathologists but generations have a tendency to be resilient and adapt.

 

[Daddydoc]

I think it will turn into tools to make us better and more efficient at our jobs, not replace us.

 

[basket687]

I'm not. I'm also not able to disclose exactly why, or who I work for, but I can say that I've developed an appreciation for the unique contextual visual recognition skills I've developed as a mediocre pathologist, while training deep learning algorithms as an undercompensated side hustle.

I've never been able to beat the computer at chess. However, very good chess players still beat computers.

Pathology residents seem to learn more quickly than deep learning algorithms.

I have a decent background in computer chess. The best computer chess programs are way way better than the best humans. Magnus Carlsen (the current world chess champion) won't stand a chance against the best chess programs. He would be lucky to even get 2-3 draws in a 10-game match.

I think computers will be able to learn simple pattern recognition in histology but they won't be able to put the patterns in context or synthesize an intelligent path report, especially for the more complex cases.

 

[Deucedano]

Molecular testing is more of a threat to replacing us than AI. The concept is interesting in theory, but that may not translate to a broad scale. A lot of these things are tested in a controlled setting with ideal conditions and people that are experts overseeing it. Your average IT person would be useless and I doubt very much hospital admin would be willing to pay someone qualified to maintain it since it would likely require a pathologist to oversee it as well. It would be much easier to replace a hospitalist with an NP or PA using AI since they are working with discrete data (lab data, vital signs, symptoms, ect). In the end someone has to take the liability which will likely require a physician. I do think it may make us more efficient requiring a smaller pathologist workforce, so future pathologists may have more trouble getting jobs.

 

[WEBB PINKERTON]

How many threads about AI does this forum need? Like Deucedano said, molecular is the threat. Future is liquid biopsies without a doubt. Bundled payments are a huge threat to.

AI is a productivity tool at best. Just look at what it has done to cytopathology. We already have AI screening devices.

Dogs sniffed out lung cancer fairly accurately the other day. And pigeons performed well looking at tissue years ago. Maybe they are both huge threats as well.

Dogs may accurately sniff out lung cancer through their super-sensitive noses, research shows

Here's another reason to love dogs. Using their super-sensitive noses, dogs were able to sniff out lung cancer in samples of human blood with 97% accuracy.

www.usatoday.com

Science | AAAS

www.sciencemag.org

 

[dr.weiner]

How many threads about AI does this forum need? Like Deucedano said, molecular is the threat. Future is liquid biopsies without a doubt. Bundled payments are a huge threat to.

AI is a productivity tool at best. Just look at what it has done to cytopathology. We already have AI screening devices.

Dogs sniffed out lung cancer fairly accurately the other day. And pigeons performed well looking at tissue years ago. Maybe they are both huge threats as well.

Dogs may accurately sniff out lung cancer through their super-sensitive noses, research shows

Here's another reason to love dogs. Using their super-sensitive noses, dogs were able to sniff out lung cancer in samples of human blood with 97% accuracy.

www.usatoday.com

Science | AAAS

www.sciencemag.org

I hear ACADEMIA is hiring sniffing dogs as pathologists to sign out cases for mere doggie treats.

FLEA PATHOLOGY NOW


How'd I do?

 

[WEBB PINKERTON]

Pigeons will be forced into client billing arrangements with the dogs for referrals.

I like FLEA PATHOLOGY NOW. That is very funny.

Thrombus we miss you.