Will automation kill off jobs in the future?

Automated systems can get hacked and shut down. Transferring every aspect of every job to AI would open companies up to having all their shit slapped, especially if it's online. Any company that pulls this shit will get hacked, especially if it puts a lot of people out of jobs.
 
The Mass Shooter Ron Soye said:
I think there has been plenty of interest in getting trucking automated, it will happen eventually, and Null is right to point out that TPTB do not want another trucker protest ever again.
Click to expand...
What a lot of people miss about things being marketed as 'automation' is that it usually- and in fact often- isn't actually autonomous AI operating things, at least not all the time.

There have already been self driving services sold which are just based on having an Indian drive your car remotely.

Tesla will never build a fully self driving car. But I bet Freightliner can sell 'self driving' trucks that drive themselves 95-99% of the time and the rest of the time the owner either hires people in their headquarters or outsources things to a 'driving service' in India to assess risks and pull over safely when they're caught in a snowstorm, need to navigate a very complex container terminal, need to interact with police officers about unusual loads at waystations, etc.
 
Mostly yes, if anything we're kneecapping what software and scripting can do to replace jobs. We have this weird system where a lot of office jobs are reliant on workers and customers/clients not knowing how to use computers very well that could collapse at any point and cause a snowball effect on other jobs. On the other hand those fields could hold indefinitely by some pyramid scheme.
 
  • Like
Reactions: Aidan
I don't want to PL here, but I know a little bit about automation that makes me skeptical that we'll ever totally replace all manual labor with automated machinery.

Automation only makes sense when the marginal cost of work produced by the previously manual process over a reasonable period of time (typically 1-5 years) exceeds the marginal cost of investing in an automation system (upkeep cost of the system + .payments for loans). Most automation systems are engineered to the application and are typically inflexible - trying to engineer automation systems for flexibility ratchets up the cost beyond what most capital holders are willing to pay, and may also impede performance depending on the application. Think about this: if you're packaging soup cans, automation makes a ton of sense since your soup packaging machine can do the work of multiple people,, so your marginal cost of going with automation versus is very low. If you're running a restaurant with a huge menu, maybe not.

The cost of the automation system is also directly proportional to the complexity of the application. If you look at a modern automotive production line, the assembly of the interior is still done by hand. Sure, the operators have a lot of help through computerized power drills and pneumatic lifts, but a person has to maneuver around the interior of this car putting all of the little bits together. Trying to get a robotic arm to do this would be outrageously expensive and complex, because chances are good you can't just program where everything in the interior goes. It's a learning process people can learn pretty quickly, whereas teaching it to a computer would take tons of trial and error data sets, and nobody wants to invest in that.

Unlike people, automation systems have tons of controllers and sensors that may require maintenance, and somebody who likely isn't a machine themselves will need to maintain it. Until we develop universal humanoid robot that can be taught to carry out any maintenance task, this will not change.
 
  • Agree
  • DRINK!
Reactions: Maggots on a Train v2, Dread First, teriyakiburns and 2 others
Soft automation is happening that is killing jobs in invisible ways all the time.

Lets say some place like Walmart makes a checkout system that is now twice as fast or more reliably handles deliveries when ordering on their website. They now have less need for checkout people and people to fix bad deliveries. They dont fire checkout or order managers but they dont hire new ones (even temp ones) like they used to for stuff like Black Friday and maybe dont replace their spots when they leave more organically.

You dont get big dramatic layoffs and firings just slow inevitable attrition. It happens in coding and dev too. Projects that used to demand 100 plus devs get handled by 30 plus instead.
 
Unironically pre-AI machine learning can and will kill off a lot of creative and white-collar jobs, moreso than any automation that requires a digital-physical interface.

One of the amazing things that’s appeared in the last few years is AI art generators, which is almost an arrow to the heart of art.

Imagine instead of paying for an artist, you can just type in some variables, upload some precedents, and textures into a generator and it’ll spit out 50 unique DONUT STEEL characters for you?
 
  • Agree
  • Dislike
  • Horrifying
Reactions: DuckSucker, Clown College, teriyakiburns and 1 other person
My current job could 100% be automated. Literally no reason to have people do it, but I know it'll never be so because a majority of our clientele is too stupid or old to even use the internet let alone a basic kiosk.

I think that's how a lot of places are. It could be easily automated but your average consumer is too ill tempered, old, or just stupid to actually make use of any interface a store/facility might have.
 
Automation will kill off certain industries, but it will take a long while to get there. Back in the very early 2000's; railroads were trying to cut down on how many people they can stick on a train and started testing these radio control beltpacks. Long story short, even with radio towers giving coverage to the entire area and in a controlled environment of a train yard, they weren't that good. If you can't get them to work in the yard, there's no way to get them to be reliable going the hundreds of miles between depots or be able to drop off certain cars in remote places as that's currently required. While trains are one thing, the idea applies to logistics as a whole, we are nowhere close to being able to automate vehicles to put around town in, there's no way in hell we're ready to automate big vehicles hauling thousands of pounds of goods. One of the big problems is infrastructure, the amount of infrastructure to require vehicles to be able to operate independently in a safe manner is immense, and no one will want to front the cost for it, and without some sort of super development in the future, I don't think it's feasable or cost effective; especially when you consider repair and maintenance to keep the entire thing working. And even if it was feasable, if something breaks, you'll always need someone to push the button or take the wheel if/when the system fails; even if the person is just sitting there waiting for something bad to happen, you'll need a trained person there to do the work the machine failed to do.

I could see retail for non-perishable goods becoming entirely robotic though; again, maybe with a person behind the counter in case the robot breaks. Customer enters the store and approaches the counter, pushes buttons on a touch screen, robot is programmed to know where everything is, brings it out on a conveyer, person swipes their card, a hatch opens, person takes stuff and then leaves. I can see stuff like that happening, but things that are perishable will need to be inspected and thrown out, which is a whole new level of difficult for a robot. Changes will come, but they are so far in the future, we shouldn't be contemplating how it will happen, but recognizing that to not get left behind, you should learn electrical work, engineering, or IT, as you'll be required to keep this shit running.
 
You must log in or register to reply here.
Back