![]() |
|
|
The above list shows replies to the following message: |
|
Msg. 08260 of 10975
(This msg. is a reply to 08248 by scubavol) |
scubavol, Good question! I confess I was caught a little flat footed when I went looking specifically for jobs already impacted and found, almost exclusively, stories about jobs TO BE impacted in the next 5-10 years.
But here is a partial list from one article;
Jobs That AI Already Replaced: Notice these are pretty much all low level "knowledge worker" jobs. That is partly because existing systems were trained in mostly confined niches where the type of work tended to have tight boundaries and was judged mostly by written or vocal output.
What I can speak to as job losses by people I know is: entry level programming and people who have jobs writing, in almost any capacity. And I can also tell you, as a learner, that there is little to no future for a "language teacher", because the AI tools are so good they are already better than native speakers in most regards.
Duolingo just dared to announce they are replacing most of their staff with AI for generating new learning content: http://www.newindianexpress.com/business/2025/May/12/ibm-replaces-about-200-hr-roles-with-ai-agents But there is also the major, VALID fear of giving these AI systems PHYSICAL movement too early. Just wait until the robotaxies really roll out and become commonplace. That is starting this year: http://www.aol.com/news/uber-ceo-says-waymo-robotaxis-125257500.html. This thing is coming on like a tidal wave. It will take a few years for word of mouth to spread successful usage from "early adopters" to companies which are more traditionally managed. And there is also an acceptance by customers delay which we have seen before with most other technological changes. Have you been mostly waited on by a robot in a restaurant, yet? I have. And the REAL AI systems are still in fairly early rollout and still undergoing a lot of careful management, because REAL AI systems don't have an "algorithm" which can be checked for logic errors. That means, they can't be assured to be safe. But soliders aren't meant to be safe, so replacing soliders with robots can be expected to part of regular warfare within the next year or three. It is just an issue if the countries involved in the war have access to the tech. Nobody can actually say HOW a REAL AI system reasons because, as I said, there are no coded alorithms. The machines are *trained* effectively with reinforcemnet when they get the right answer or give the right behavior. Robots at Google taught THEMSELVES to play soccer more than two years ago: So, not as complete and robust an answer as I thought I could give you. But it is coming. And believe me, if robots two years ago could teach themselves to play soccer, it isn't that hard to predict that fairly crude physical tasks will be displaced fairly rapidly given the exponential rate of improvement in these systems. One big thing which will delay rollout is acceptance of the population to dealing with an AI. Expect a lot of "bigotry" for the first decade or so.
|
|
|
|
|
© Webpage Design Copyright 2003-2011 http://www.atomicbobs.com/
|