The General Economics Thread

Manufacturing? Sure. Accounting? Maybe. But machines properly replacing service jobs? Maybe in fast food places but in literally every other service job that's a big no. Robots are still too dumb.

Also, if memory serves, service robots have already been used but they tend to receive less positive responses by customers than human service.

It's kinda silly but at the local Mc D where they have the order-taking machines, instead of some of the staff being at the counter taking order, they have the staff by the machines teaching people how to use the machines....

You mean like so-called "self-service tills" at supermarkets? Then again, there is somewhere where machines can take over the work of humans — online shopping...

The last time I went shopping there was a lady at the next self-checkout over screaming "BUT WHAT DOES IT MEAN!?" over and over at the top of her lungs while the machine was asking her how she wanted to pay. The screen had huge icons showing that the options were debit, credit or cash. She was maybe 30. It turned out that she wanted to pay with cash. She knew how to put cash in, but she had no idea about which option to pick to do that.

There will always need to be at least a few employees because people will convince themselves they don't know how to do something that is actually really simple, or blindingly obvious.

True but you can still often reduce the number of people needed. 1 person can watch over 6 self service tills, instead of having 4 human tills with 4 people manning them.

The throughput of those self-service checkout systems is lower than a regular line, though, partly because of limited space making it exceedingly awkward to check out a large order, but also because the system attempts to prevent theft by making you scan and bag each item individually.

Couple that with customers being less practiced at the activities involved in checking out, often requiring assistance, and those systems end up having significant drawbacks.

We won't get fully automated checkouts in grocery stores without making much wider use of RFID technology. Better to dispense with storefronts entirely and make the entire shopping experience based on online order/delivery.

edited 28th Nov '15 10:32:47 AM by Fighteer

Thing is you don't need an exact till for till output equality. You just need the number of self service tills supervised by one person (say 6) to be faster then if that one employee was running a normal till.

But if individual customers have a poorer experience checking out, you haven't actually saved anything except the number of meat sacks on your payroll. Those self-service tills only work for small orders; you still need regular checkout lanes for anyone who buys a significant amount of stuff.

edited 28th Nov '15 10:56:09 AM by Fighteer

The point is that it will happen slower than the futurist doomsayers think, but it will still happen, and arguably has been happening as part of the dilution of manufacturing.

Fighteer is right in that businesses and government are slower to adapt internally, but external adaptations, in the consumer market, are very quick. While 911 might still rely on archaic analog telephone systems, the consumer end is unrecognizable. But the inertia of business and government systems will stand in the way of efficiency-ushered doom.

Automation in manufacturing is here and will only get bigger. Automation in office environments — particularly in government and heavily regulated industries — is incredibly slow. Automation in service industries has major consumer adoption hurdles to overcome.

Home automation is largely out of the reach of the majority of consumers who aren't rich enough to afford it.

edited 28th Nov '15 10:58:41 AM by Fighteer

And full automation in the military will be the slowest of them all, especially where Nuclear Weapons are involved.

Although, maybe upgrading the nuclear weapon security systems to something developed in the last 40 years might be a good idea. Yeah, the current system is practically impossible to crack, but it might well break completely.

I'm totally up for uncrackable nukes that wouldn't even fire if they were hacked.

With Blue Circle warheads?

Automation will eliminate many jobs, and create more than they eliminate. Always have, always will.

I don't see it, myself. We could easily reach a point where computers/robots can do 90% of anything we could possibly want done better than we ever could. Problem solving tasks, physics, computer programming, data entry, manufacturing, machines can do a lot of this stuff better than we can.

Well, if we reach the point where automation can literally replace most people, then we'll have reached what's called a "post-work society". At that point, we'll essentially have to implement some sort of highly-redistributive economic system where taxes on the wealthy (presumably the people who own all the automation that's replacing workers) pay for a basic income that allows everyone else to survive. If you don't do that, you get people literally starving on the streets, and when that sort of thing happens en masse you end up with a revolution — the torches-and-pitchforks kind.

Of course, this is all far enough in the future that it's essentially science fiction. In the meantime, the most we can look forward to is disruptive technologies causing problems in certain industries. Self-driving cars putting all the truck drivers out of work sort of thing, which is no different than things that have happened before (electric lights putting candle-makers out of business, etc).

edited 28th Nov '15 8:40:25 PM by NativeJovian

And there are still candle companies and makers out there, for various different kinds and purposes (from domestic aesthetic to religious purposes), in any case.

edited 28th Nov '15 8:42:43 PM by Quag15

Re: Automation in Office Environments

Here's some snippets from the book.

Writing

Computer Maintenance

Data Analysis

How are things looking for recent graduates

edited 28th Nov '15 9:57:31 PM by probablyinsane

Apparently Volvo are working on self-driving trucks and buses.

Yeah, well...

The loom, the printing press, the steam engine, the tractor and many other sorts of machines replaced what would be hundreds of tailors, scribes, laborers, farmers and workers, yet the employment increased, people doing those jobs went on to do something else.

While these things make jobs easier, they can't fully replace them. For a computer to assemble an article someone has to write and document the content first, a administrator can manage thousands of computers but you still need human hands to repair and configure them, those data analysis probably mean that those other 4 people aren't out of a job but assigned to do something else. Which is what usually happen.

Automation either allows you to do something else with the same workforce or allows you to get bigger without massive costs.

The rest is not related to automation as much it is related with the screw ups of the US and the global market, specially how shitty the work conditions are in the US. The US and Europe also no longer grow fast enough to absorb their students, meanwhile emergent countries with booming economies are having a deficit of skilled STEM students.

Stagnated economies and little growth screw graduates and the average worker more than automation ever will.

The lingering problem with automation replacing everyone is that today there are companies that realized that you need to pay a decent wage and provide decent jobs otherwise you will have no customers for the stuff you make.

By the time automation replaces most if not all work, we will no longer be playing by the current economic models of today, we'll probably work with something that can be describe with a pre-post scarcity economy, where work is no longer scarce but resources may still be.

Again, I'm worried about security-bots.

Thankfully, robots will suck at detecting allegiance for a long time.

Are you really sure?