4 ms·

Med school is probably all right. Of course robotics and advancements in AI are going to take away some of these jobs (personally I'd prefer IBM Watson to look

Med school is probably all right. Of course robotics and advancements in AI are going to take away some of these jobs (personally I'd prefer IBM Watson to look at my genetic data and results of the blood tests, rather than some human doctor), but there would certainly be some jobs in medicine.

rohern 14y ago
I love this comment. This is exactly how parents need to be thinking. Yes, being a doctor is a relatively safe career choice now, but your kid is going to be graduating from college in a decade...
- rayiner 14y ago
  A decade ago, I was posting on web forums looking exactly like HN on a 1.9 GHz P4 laptop. My laptop has gotten smaller since then, and my cell phone faster, but the world doesn't look that different.
  - dchichkov 14y ago
    A decade ago a consideration that a computer can diagnose a patiant better than doctor was a joke. Now check out this: https://www.youtube.com/watch?v=HZsPc0h_mtM https://www.youtube.com/watch?v=HZsPc0h_mtM
    - rayiner 14y ago
      Expert systems of this kind have existed for decades now, and what Watson does could have been done, on a smaller scale commensurate with CPU/memory power of the time, in the 1990's.
      - dchichkov 14y ago
        It is a question not of system existence, but of performance relative to your average doctor. And once commercially deployed systems would start outperforming your average doctors these high-paying doctor jobs are bound to change somehow. The answer to that question was a definite no, some 20 years back. And no, some 10 years back. Now, the answer is yes. And in some 20 years it will be resounding yes. Why? First, because medicine gets less and less 'shaman based' and more and more evidence based. Some 20 years back you could't scan a region of your DNA and figure out that you have a particular versions of couple of proteins that wrecking havoc in your particular case. Now you can. Second, there've been improvements in statistical models and natural language processing. To the level that computers can actually outperform humans in these murky areas. An IBM success with Jeopardy is a good example. Combine these two, take into account that computers are less likely to overlook hard evidence (like DNA data, incopatibilities with medications, etc), and you would also want an expert system to look at your health data, diagnose, recommend treatments and monitor the progress. Either way. I still think that a "med school is probably Ok" could be a good advice to a kid.
        rayiner 14y ago
        I understand that this is HN and people are optimistic about the potential of technology, but betting against technology changing everything tends to be a winning bet more often than the otherwise. When I was a kid, I wanted to be an aerospace engineer. I read books in the 80's talking about how we'd have manned Mars missions in 25 years. I actually went to engineering school for aerospace, but by then I had realized humans wouldn't go into space in a real way within my career. Is it possible that some derivative of Watson is going to replace doctors? Maybe. Or maybe it'll just change what doctors do for a living in the same way auto-pilot changed what pilots do for a living. My bet is that 20 years from now isn't going to look that dramatically different from 20 years ago, in a fundamental way. I remember 1993 (a year before President Clinton's first term and a year before Amazon was founded) and it wasn't all that different. We made phone calls and sent faxes instead of using e-mails, there was no Twitter, Facebook, or smart phones, but life was pretty similar.
        nostrademons 14y ago
        Dude, I look at 1993 and compare it to today and am totally amazed at how much life has changed. In 1993, your circle of friends was pretty much limited to the people in your town. Maybe you had a pen pal, but that was considered to be a rare novelty and took real work to maintain. Now, I have friends all over the world, that I keep up on their day-to-day lives. When an earthquake hit SF in 1989, my reaction was "Oh, that's too bad, I feel bad for anyone who died. I'm sure we'll be talking about this in school today." When an earthquake hit Japan in 2011, my reaction was "Has anyone heard from Jennifer yet? Is she okay? I know she was in Sendai..." And I had friends all over the globe - India, Colombia, DC, Boston, Morocco, LA - asking the same thing. In 1993, if you were meeting a friend for an event at an unfamiliar place, you called them up beforehand, specified an exact time and place that you would be, wrote down the directions or brought along a road map, and prayed they would be there when they said they would. If you got lost, you had to pull over until you could look at the map and memorize new directions. If you couldn't find each other in the crowd, you were out of luck. If you hit traffic on the way in, you'd hope they didn't think they were being stood up and apologize profusely. Now, your calendar pops open a notification when it's time to leave, taking into account traffic. You hit a button, and your cell phone navigates to the destination, automatically rerouting you if you take a wrong turn. You can't find your friend, you call or text them. You don't know what they look like, you call them and see whose phone rings. In 1993, if you wanted to start a company, you worked your butt off for a big employer to learn the ropes of your industry. You spread your network as far as it could go. You sought out mentors. You begged for funding. And then you hired lots of people to make your dreams come true, and if your dreams were wrong, they all got laid off a couple years later. Now, you Google [startup]. You read a bunch of blogs. You download lots of open-source software. You build something, you launch it, you post it on Hacker News and a bunch of blogs, and if your dreams were wrong, you've wasted maybe a year of your life. If your life isn't all that different from 1993, it's more likely because you're still living in the past than because the world is the same as 1993. "The future is here, it's just not evenly distributed yet." For some people, the world is wildly different from what they expected 20 years ago.
rayiner 14y ago
Betting against AI has been a winning play since the history of AI. AI isn't any closer to replacing a human doctor today than it was in the 1980's. Maybe some day, but we're talking fundamental breakthroughs in the field, not incremental advances from existing techniques.
- redschell 14y ago
  True, but look at the timeline for your daughter. Assuming she was recently born, and, given that few from a school like Penn would want to be a rural family physician, assuming she's going into an elite specialty (cardiology, neurology, oncology, etc), she's THREE DECADES away from getting started in her line of work. And that's a generous projection. More likely, she'll need to be hyperspecialized, an expert in some obscure niche area like cancers of the eye, in which case she would need more and more education, and I doubt Google's engineers are going to just chill out and wait for her. Not trying to be a downer here. I'd rather the world have more aspiring doctors than talent agents, but let's also respect the march of progress, and acknowledge the cruel fate of all that stands in its way.
  - rayiner 14y ago
    I understand the point your making, and I don't disagree with you, but let's say I'm skeptical of Google's ability to do anything other than help my find my porn more easily. AI has just been such a miserable failure for decades. Indeed, Google's big advancement came from incorporating human intelligence (encoding in the graph of links between pages) into the search process! I'm not saying it's not going to happen, but rather that it's like manned missions to mars. Always just 20 years away... As an aside, being a rural doctor is amazing. They're in very short supply, and therefore the salaries are correspondingly quite high.
- redwood 14y ago
  "since the history of AI" <-- lol. This could be said about any recent technology with a short history.... electronics in your pocket, for example
  - 14y ago
larrys 14y ago
"AI are going to take away some of these jobs" Population aging and people will always abuse their body and create a need for medical care delivered by a human. And as people live to older ages the base individual becomes more likely to need medical care (wish I had a better way to put this) whereas in years past that person would have died at an earlier age and not required any medical care at all.

rohern 14y ago

I love this comment. This is exactly how parents need to be thinking. Yes, being a doctor is a relatively safe career choice now, but your kid is going to be graduating from college in a decade...

rayiner 14y ago
A decade ago, I was posting on web forums looking exactly like HN on a 1.9 GHz P4 laptop. My laptop has gotten smaller since then, and my cell phone faster, but the world doesn't look that different.
- dchichkov 14y ago
  A decade ago a consideration that a computer can diagnose a patiant better than doctor was a joke. Now check out this: https://www.youtube.com/watch?v=HZsPc0h_mtM https://www.youtube.com/watch?v=HZsPc0h_mtM
  - rayiner 14y ago
    Expert systems of this kind have existed for decades now, and what Watson does could have been done, on a smaller scale commensurate with CPU/memory power of the time, in the 1990's.
    - dchichkov 14y ago
      It is a question not of system existence, but of performance relative to your average doctor. And once commercially deployed systems would start outperforming your average doctors these high-paying doctor jobs are bound to change somehow. The answer to that question was a definite no, some 20 years back. And no, some 10 years back. Now, the answer is yes. And in some 20 years it will be resounding yes. Why? First, because medicine gets less and less 'shaman based' and more and more evidence based. Some 20 years back you could't scan a region of your DNA and figure out that you have a particular versions of couple of proteins that wrecking havoc in your particular case. Now you can. Second, there've been improvements in statistical models and natural language processing. To the level that computers can actually outperform humans in these murky areas. An IBM success with Jeopardy is a good example. Combine these two, take into account that computers are less likely to overlook hard evidence (like DNA data, incopatibilities with medications, etc), and you would also want an expert system to look at your health data, diagnose, recommend treatments and monitor the progress. Either way. I still think that a "med school is probably Ok" could be a good advice to a kid.
      - rayiner 14y ago
        I understand that this is HN and people are optimistic about the potential of technology, but betting against technology changing everything tends to be a winning bet more often than the otherwise. When I was a kid, I wanted to be an aerospace engineer. I read books in the 80's talking about how we'd have manned Mars missions in 25 years. I actually went to engineering school for aerospace, but by then I had realized humans wouldn't go into space in a real way within my career. Is it possible that some derivative of Watson is going to replace doctors? Maybe. Or maybe it'll just change what doctors do for a living in the same way auto-pilot changed what pilots do for a living. My bet is that 20 years from now isn't going to look that dramatically different from 20 years ago, in a fundamental way. I remember 1993 (a year before President Clinton's first term and a year before Amazon was founded) and it wasn't all that different. We made phone calls and sent faxes instead of using e-mails, there was no Twitter, Facebook, or smart phones, but life was pretty similar.
        nostrademons 14y ago
        Dude, I look at 1993 and compare it to today and am totally amazed at how much life has changed. In 1993, your circle of friends was pretty much limited to the people in your town. Maybe you had a pen pal, but that was considered to be a rare novelty and took real work to maintain. Now, I have friends all over the world, that I keep up on their day-to-day lives. When an earthquake hit SF in 1989, my reaction was "Oh, that's too bad, I feel bad for anyone who died. I'm sure we'll be talking about this in school today." When an earthquake hit Japan in 2011, my reaction was "Has anyone heard from Jennifer yet? Is she okay? I know she was in Sendai..." And I had friends all over the globe - India, Colombia, DC, Boston, Morocco, LA - asking the same thing. In 1993, if you were meeting a friend for an event at an unfamiliar place, you called them up beforehand, specified an exact time and place that you would be, wrote down the directions or brought along a road map, and prayed they would be there when they said they would. If you got lost, you had to pull over until you could look at the map and memorize new directions. If you couldn't find each other in the crowd, you were out of luck. If you hit traffic on the way in, you'd hope they didn't think they were being stood up and apologize profusely. Now, your calendar pops open a notification when it's time to leave, taking into account traffic. You hit a button, and your cell phone navigates to the destination, automatically rerouting you if you take a wrong turn. You can't find your friend, you call or text them. You don't know what they look like, you call them and see whose phone rings. In 1993, if you wanted to start a company, you worked your butt off for a big employer to learn the ropes of your industry. You spread your network as far as it could go. You sought out mentors. You begged for funding. And then you hired lots of people to make your dreams come true, and if your dreams were wrong, they all got laid off a couple years later. Now, you Google [startup]. You read a bunch of blogs. You download lots of open-source software. You build something, you launch it, you post it on Hacker News and a bunch of blogs, and if your dreams were wrong, you've wasted maybe a year of your life. If your life isn't all that different from 1993, it's more likely because you're still living in the past than because the world is the same as 1993. "The future is here, it's just not evenly distributed yet." For some people, the world is wildly different from what they expected 20 years ago.

rayiner 14y ago

Betting against AI has been a winning play since the history of AI. AI isn't any closer to replacing a human doctor today than it was in the 1980's. Maybe some day, but we're talking fundamental breakthroughs in the field, not incremental advances from existing techniques.

redschell 14y ago
True, but look at the timeline for your daughter. Assuming she was recently born, and, given that few from a school like Penn would want to be a rural family physician, assuming she's going into an elite specialty (cardiology, neurology, oncology, etc), she's THREE DECADES away from getting started in her line of work. And that's a generous projection. More likely, she'll need to be hyperspecialized, an expert in some obscure niche area like cancers of the eye, in which case she would need more and more education, and I doubt Google's engineers are going to just chill out and wait for her. Not trying to be a downer here. I'd rather the world have more aspiring doctors than talent agents, but let's also respect the march of progress, and acknowledge the cruel fate of all that stands in its way.
- rayiner 14y ago
  I understand the point your making, and I don't disagree with you, but let's say I'm skeptical of Google's ability to do anything other than help my find my porn more easily. AI has just been such a miserable failure for decades. Indeed, Google's big advancement came from incorporating human intelligence (encoding in the graph of links between pages) into the search process! I'm not saying it's not going to happen, but rather that it's like manned missions to mars. Always just 20 years away... As an aside, being a rural doctor is amazing. They're in very short supply, and therefore the salaries are correspondingly quite high.
redwood 14y ago
"since the history of AI" <-- lol. This could be said about any recent technology with a short history.... electronics in your pocket, for example
- 14y ago

larrys 14y ago

"AI are going to take away some of these jobs" Population aging and people will always abuse their body and create a need for medical care delivered by a human. And as people live to older ages the base individual becomes more likely to need medical care (wish I had a better way to put this) whereas in years past that person would have died at an earlier age and not required any medical care at all.