Amazon rolls out machines that pack orders and replace jobs • Hacker News Board

Re: Amazon rolls out machines that pack orders and replace jobs

#381

Post by bschne » Mon, May 13, 2019, 7:00 PM UTC

Animats wrote: post #358

This is a minor improvement in packaging. It's not automated picking, which will be a big deal if Amazon cracks that. They've been trying and failing for years now. Amazon gave up on their robotic picking challenge after 2017. Here's the winner of the last challenge.[1] Way too slow and not reliable enough. Amazon has about 100,000 robots now. Most are Kiva machines, moving storage racks to human pickers. Here's what…

Here's how one of the biggest online electronics retailers in Switzerland handles storage / logistics for their smaller items [1].

As is the case with your amazon example, putting the items in the storage containers and picking them for shipping is still handled manually. I'm really curious as to how this compares to the moving shelf idea efficiency-wise and how a possible "solution" (if it is a problem that needs solving...) for the picking could look like, e.g. standardized containers for small items that could be handled by a machine or something like that.

[1] https://www.youtube.com/watch?v=Scql9_7RjfU

Re: Amazon rolls out machines that pack orders and replace jobs

#382

Post by stale2002 » Mon, May 13, 2019, 7:01 PM UTC

nerdponx wrote: post #327

o_p wrote: post #312

Earlier quoted context omitted.

>machines bring lower prices Do you have any source about this? If the costumer is already willing to pay a price there would be no point in lowering the price just because you automated the process. Or better said, theres no point in investing in machines if your profit margin is not going to get wider.

It's worth expanding on this. Because I think it gets lost in the rhetoric, whether deliberately or by accident. A profit-maximizing firm sets prices that maximize its expected profit. This seems tautological, but it's important and I think it's implications aren't appreciated. If your prices are already beating the competition, and you already have a solid reputation for low prices, there is no incentive to continue…

> there is no incentive to continue lowering prices,

Yes there is. Lower prices means that consumers will buy more.

Unless you are going to argue that the elasticity of demand is perfectly inelastic. Such cases are rare.

Re: Amazon rolls out machines that pack orders and replace jobs

#383

Post by yomly » Mon, May 13, 2019, 7:05 PM UTC

elamje wrote: post #301

Articles about Amazon are quick to mention the workplace conditions or robots automating something. Rarely do they mention that machines equal lower prices for the consumers 9 times out of 10. You can think what you want about Amazon but I’m not complaining that I have 100’s of options for my phone case, all for 1/5 the price of the previous days buying them at Verizon or Apple. Edit: would be awesome to hear about A…

When you collect as much data as Amazon, you can start attributing incremental worker time spent picking/carrying down to the SKU level. This means that cost of worker time is less for things like books (a picker can carry many and move quickly) while a 60" TV might require two pickers moving slowly.

Better attribution of cost means more margin to play with and cost savings that can be pushed onto customers in a fairer sense - especially if your pricing is automated.

Interestingly Kiva may then selectively reduce the costs associated with bulky goods as they're probably better at moving big stuff and travel faster/shorter.

Re: Amazon rolls out machines that pack orders and replace jobs

#384

Post by shiburizu » Mon, May 13, 2019, 7:11 PM UTC

dzmien wrote: post #350

smpetrey wrote: post #303

Anyone parading around like this is bad news has never worked in an assembly line or warehouse job. This might reduce the workforce overall long-term, but honestly this is hard back-breaking long-workday labor. Anything that improves the livelihood of the warehouse workers is a win in my book.

I worked for Amazon in what was then a brand new warehouse for a few months in 2015. I worked 10 hour shifts, from 20:00 to 07:00 (the extra hour was for lunch and breaks). My job was picking orders. It seemed easy at first (and it was, but mentally, not so much). All I had to do was stand in one place while robots brought me pod after endless pod of products. There was nothing to it, and that was the problem. The po…

I'm glad you made that decision for yourself and I think such perspectives are missing from the discourse about working-class folks dealing with addiction.

Re: Amazon rolls out machines that pack orders and replace jobs

#385

Post by logicprog » Mon, May 13, 2019, 7:11 PM UTC

UnFleshedOne wrote: post #324

logicprog wrote: post #144

Earlier quoted context omitted.

I one hundred percent agree with you on what the results would be-- without there being repercussions for not persevering with anything, and with all bad choices safely having the corners rounded off by UBI, most people won't do anything much. And as much as they feel like they don't want to do work now , this lack of any motivation from outside themselves will have a deleterious effect on people. I mean, picture wha…

The problem with not using free time wisely is that there is usually supply and fragmentation issues -- not enough free time (or at least feels like it) and chunks that feel small enough that you have trouble fitting anything significant in before you have to drop everything and go make another buck. So the time is spent on useless things that are easy to pick up. For example, unless currently possessed by a Strange…

Okay that's a good point. Also I love the DF reference, DF is my all-time favorite game.

My original comment somewhat lacked empathy, perhaps because I don't suffer from problems with free time fragmentation because I'm always possessed by a Strange Mood. I still think my point stands though. I meet a lot of people with what I know to be fairly large chunks of time that don't do anything with them anyway, not because they feel like they can't achieve something in that time but because they don't have any hobbies or ideas, or particular skills. In fact, that's true for most people I meet, and it's actually difficult for me to tell whether that's an artifact of perceived free time fragmentation or not. And this is in a culture where people are 100% encouraged to develop skills and hobbies.

But think about it, as much as we hear about retired people doing epic things, we also don't hear about retired people, like my grandparents, being terminally bored and ending up back to the workforce.

Here's the thing. As much as jobs can be boring and soul-sucking, to an extent they provide more of something then doing nothing while on UBI does. In addition, they provide a lot of training for life, like perseverance through boring things, and people skills (this is my experience at least) that you can't really get anywhere else. Plus, while jobs might be unfun, developing any kind of skill and competency at something genuinely gives people a sense of self-worth. Often, jobs provide the sort of skill that is easier to master than, say, art or poetry. Unless UBI comes with some sort of education or structured skill development, it really won't be good for people in the way you're suggesting.

I'm willing to consider it possible however that there might be more people capable of learning skills and achieving self-actualization than we currently see manifest in society, and that with UBI those people will be freed up to do things. I just don't think that it's a totally positive thing. Besides all kinds of other logistical issues and issues like it ending up subsidizing drug addictions and such.

Re: Amazon rolls out machines that pack orders and replace jobs

#386

Post by celtain » Mon, May 13, 2019, 7:12 PM UTC

nerdponx wrote: post #327

o_p wrote: post #312

Earlier quoted context omitted.

>machines bring lower prices Do you have any source about this? If the costumer is already willing to pay a price there would be no point in lowering the price just because you automated the process. Or better said, theres no point in investing in machines if your profit margin is not going to get wider.

It's worth expanding on this. Because I think it gets lost in the rhetoric, whether deliberately or by accident. A profit-maximizing firm sets prices that maximize its expected profit. This seems tautological, but it's important and I think it's implications aren't appreciated. If your prices are already beating the competition, and you already have a solid reputation for low prices, there is no incentive to continue…

>If your prices are already beating the competition, and you already have a solid reputation for low prices, there is no incentive to continue lowering prices, even as you cut costs.

If they could reach many more customers by lowering prices, even a monopoly could potentially find it profitable to continue lowering prices as they cut costs.

A monopoly could also continue to cut costs to head off the threat of antitrust litigation.

Re: Amazon rolls out machines that pack orders and replace jobs

#387

Post by blairanderson » Mon, May 13, 2019, 7:13 PM UTC

mschuster91 wrote: post #218

blairanderson wrote: post #168

Earlier quoted context omitted.

You mentioned two processes that have been in place with Amazon Vendors/Manufacturers for many years. 1. Amazon Direct Fulfillment (dropship): Order is shipped directly from the manufacturer to the customer. Amazon pays for the shipping label and contractually "own" the item once it leaves the MFR doors. Amazon incentivizes Manufacturers by paying the freight and loosening shipment deadlines. [1] 2. Ships In Own Cont…

> Ships In Own Container (Amazon SIOC) certification puts the responsibility on the manufacturer to certify their products can be shipped without extra packaging. Sorry, but how can this be a good idea? I'd hate if I ordered something expensive, for example a digital camera, and every porch pirate (or shady parcel delivery subcontractor) immediately sees "this thing is worth a couple grand", compared with the usual A…

Its usually a brown cardboard overbox with a barcode that has extensive droptesting.

More common these days is for manufacturers to produce items, and separately purchase very durable over-boxes that are sent to eCommerce customers like amazon/walmart/bestbuy/etc.

Re: Amazon rolls out machines that pack orders and replace jobs

#388

Post by georgespencer » Mon, May 13, 2019, 7:20 PM UTC

pkaye wrote: post #379

gdy wrote: post #369

Earlier quoted context omitted.

"And for $13.75 an hour, it definitely isn't worth it." A billion of people around the world would be happy to work with that hourly rate.

Not at the US cost of living. Nor would those billions of people be paid a that rate if their cost of living is lower than the US.

I think it's quite likely, but unknowable, that folks would accept the income:cost of living ratio of the US if it came with the associated sociological benefits. I think you're underestimating how disproportionately impoverished many people around the world are.

Re: Amazon rolls out machines that pack orders and replace jobs

#389

Post by yhoneycomb » Mon, May 13, 2019, 7:26 PM UTC

dzmien wrote: post #350

smpetrey wrote: post #303

Anyone parading around like this is bad news has never worked in an assembly line or warehouse job. This might reduce the workforce overall long-term, but honestly this is hard back-breaking long-workday labor. Anything that improves the livelihood of the warehouse workers is a win in my book.

I worked for Amazon in what was then a brand new warehouse for a few months in 2015. I worked 10 hour shifts, from 20:00 to 07:00 (the extra hour was for lunch and breaks). My job was picking orders. It seemed easy at first (and it was, but mentally, not so much). All I had to do was stand in one place while robots brought me pod after endless pod of products. There was nothing to it, and that was the problem. The po…

> I was literally just a robot arm with a human brain attached to it, and for what?

This is how I felt as an undergrad working in a research lab. Endless, annoying pipetting. Labeling tubes. Bitchwork that the postdoc didn't want to do. I kept thinking... there's no way they can't come up with a robot to do this.

My guess is the technology to do it is there but the motivation isn't - most undergrads (me included) work in research labs for free with the hope of getting into med school.

Luckily, I'm in med school now so I don't have to worry about being taken advantage of in that way. But every time I think about how the system does this to so many premeds, it pisses me off.

Re: Amazon rolls out machines that pack orders and replace jobs

#390

Post by irrational » Mon, May 13, 2019, 7:38 PM UTC

ekianjo wrote: post #186

CodeMage wrote: post #121

Earlier quoted context omitted.

> twice the speed = twice less developers in a blunt way Except that's not how that works. Increasing the productivity of a job does not eliminate it. A job is eliminated when it's no longer necessary. Unless you live in a parallel dimension, we're both in the world where there's more and more software development being done every day. Twice the speed results in twice the output, not half the developers.

> Increasing the productivity of a job does not eliminate it. Of course it does. Not on an individual level, but if suddenly one employee can do three times more work, there is no need to recruit 2 more full time jobs to address the increasing needs. So on the whole, productivity will destroy jobs, but the good news is that the additional returns will create other jobs in other areas.

We have way more work than we have time for (or the budget to hire that many more developers). If automation allowed each of us to do 3x more work, it probably still wouldn't be enough.