AI & Jobs
- Thread starter cork
- Start date
Covid with a twist
LLMs can't nativly control anything, however decision making in control systems could be trusted to the output of AI systems. That's something that will be more of a concern with more advanced AI models, but they're some way off yet.
There have been warnings from very senior people on the ultimate risk of AI when it self learns. So there will be the need for very strict controls. And probably an extreme doubling down of laws on use of personal data without permission.
In terms of jobs, the jobs will move to more skilled workers who design the AI replacements. This may be an issue for general workers but I can see multiple jobs still requiring staff. And there will be shorter working weeks, more leisure time and a larger service industry.
Looking at the areas where I work, manufacturing is still a long way from being run by AI, and the solutions being introduced are very expensive, expensive to maintain but deliver good results. The labour moves to design, implementation, control, and review. The technology needs to get a lot cheaper for wider implementations. Sure, you can get robotic packers which remove people from a packing line, but change your pack format and you are back to square 1 with a lot of redesign for the robot. The ultimate aim being the robot would redesign itself.
I have seen situations where new plant machinery is remotely monitored for performance, gets a glitch and shuts down. The fitters on the plant can’t go in and hit it with a hammer any more to get it rebooted. The equipment manufacturer sends someone in to fix, and the downtime goes from hours to day and the plant is one a shut down due to a computer monitoring in another country.
In terms of jobs, the jobs will move to more skilled workers who design the AI replacements. This may be an issue for general workers but I can see multiple jobs still requiring staff. And there will be shorter working weeks, more leisure time and a larger service industry.
Looking at the areas where I work, manufacturing is still a long way from being run by AI, and the solutions being introduced are very expensive, expensive to maintain but deliver good results. The labour moves to design, implementation, control, and review. The technology needs to get a lot cheaper for wider implementations. Sure, you can get robotic packers which remove people from a packing line, but change your pack format and you are back to square 1 with a lot of redesign for the robot. The ultimate aim being the robot would redesign itself.
I have seen situations where new plant machinery is remotely monitored for performance, gets a glitch and shuts down. The fitters on the plant can’t go in and hit it with a hammer any more to get it rebooted. The equipment manufacturer sends someone in to fix, and the downtime goes from hours to day and the plant is one a shut down due to a computer monitoring in another country.
RonnieShinbal88
Registered User
- Messages
- 251
A summary of forecasts from various experts below:
When will singularity happen? 1700 expert opinions of AGI [2024]
Experts believe artificial general intelligence will occur around 2050, & plausibly sooner. Explore 955 expert opinions about singularity
"The consensus view was that it would take around 50 years in 2010s. After the advancements in LLMS, some leading AI researchers updated their views. For example, Hinton believed in 2023 that it could take 5-20 years."
A minority of 21% of participants predicted that singularity will never occur.
Gordon Gekko
Registered User
- Messages
- 7,487
Ah yes, the old management consultancy conundrum.
I have three kids and can’t make ends meet…
Accenture’s advice is to get a better job.
McKinsey’s advice is to kill two of my kids.
So the stepmother in Hansel & Gretel was a partner in McKinsey all along!
Flybytheseat
Registered User
- Messages
- 224
No she was way too old. They burn out and leave by 30 in most cases.