Moving through an environment, self-healing, reproduction, and being powered by pretty much anything that grows are last on the list.
And Amazon ships all kinds of stuff in various packaing formats etc, so is kind of the worst case here. I could see a company that produces and packages its own stuff choose differently here, and design packaging specifically with its logistics robot capabilities in mind.
I think the issue is that there are some tasks that might seem easy for a robot to do but they are not, either too expensive vs human capital or the failure rate is not tolerable. The human body is a pretty amazing machine.
You would guess wrong then. Robots don't handle variability well at all. Making a robot that can pack anything like a human can is a huge challenge.
The robotics part is more like "how do I make a robot that can pick up a pen, a 50kg weight, and a silk blouse, and make it put all three in the same package without messing up?" In a factory that would be three distinct robots because that variability in weight, size, surface friction, malleability, etc makes it a hard problem. In a warehouse it would need to be one robot. Even if you compromise and accept additional packaging like putting soft items in boxes or bags it's still quite difficult.
Though, frankly, Amazon often seems to do a terrible job of optimizing packing--both over and under. I'm guessing that within some range of the right size and some range of not having too much breakage, it's cheaper for them to half-ass it to some degree.
For the shorter term though, imagine a telepresence robot so you can hire workers from anywhere in the world.
Order picking is, in the grand scheme of human employment, essentially unskilled labor. It takes a bit of practice to get efficient, and a certain level of perseverance and endurance to operate at top efficiency, but it doesn't take any form of higher study, credentialing or apprenticeship to become qualified at. Scaling seasonal workers in this role is relatively trivial in terms of expense compared to purchasing robots.
Final note: humans take years to become obsolete and replaced by newer generations. The same isn't true with robots. If you can get full use out of one year round, it is much easier to justify investing now rather than 5 years from now than if you only get two months a year use out of it.
The cost of living is not just related to salary but also energy being 3x more expensive than usual, house prices, food prices going up 30%, Brexit, etc.
Any increase in wages is taken by landlords who increase the rent. The competition for a house isn't the next door house, it's not living in the city at all.
1mil for a robot isn't that much. Does anyone make a robot that can work (at least) as fast as 4 people and never need maintaining etc?
Amazon's problem is they hit such a scale they are starting to run out of wage slaves. They either broke them already or the potential worker has friends or family broken by them so they are wise to the deal. If this wasn't the case they would have little interest in automation.
[0]https://fee.org/articles/no-slavery-did-not-make-america-ric...
But your original argument was about the economics for the enslaver, not for the population.
> It's the same economics that made slavery not viable. Sure you don't have to pay labor costs with a slave but if they break their arm you have to call a doctor [...] It's much more economical to work humans that you've paid a wage to the bone [...] wage slaves beat actual slaves every time.
You're trying to make a point about capital vs operational expenses, but your example of US slavery's economic viability is both poor and distracting. Enslaving people is generally a financial boon for the enslaver. Focusing on the economic costs of owning your laborers and discounting the lack of wages sounds a lot like the "Positive Good" arguments from the mid 1800s:
> Slavery is said to be an evil... But is no evil. On the contrary, I believe it to be the greatest of all the great blessings which a kind Providence has bestowed upon our glorious region... As a class, I say it boldly; there is not a happier, more contented race upon the face of the earth... Lightly tasked, well clothed, well fed—far better than the free laborers of any country in the world, ... their lives and persons protected by the law, all their sufferings alleviated by the kindest and most interested care....
https://en.wikipedia.org/wiki/Slavery_as_a_positive_good_in_...
Basic human labor planetwide is generally not expensive enough for full automation yet. This is not lost on an HR-heavy company like Amazon that employs over a million humans.
Such an equilibrium is a long way away too. We are systemically undervaluing the life energy of our species, en masse.
- Robots/automation
- Self-service
- Poorer service
- A lot of people collectively deciding they just just don't need $X if it's going to cost that much
Compare the US overall to many places in southeast Asia for example. There is generally a lot more personal service and staffing overall where the labor is cheaper.
A lot of people decide that it's their God-given right to enjoy a certain service at a certain price, even if it requires exploited labour to achieve.
They will actively support legislation that enables this sort of exploitation, because again, this service is too nice to pass on.
I really don't think consumers like automation. There are some cases where's it's convenient (vending machine) and times where it isn't (dining in). But it's being forced en mass because vendors can get away with it, not because there's demand but because suppliers like it.
Either way I'd personally pay more to use an app, so win-win for me.
This should be illegal as it really discriminates against elderly. My 75 year old dad has an old shitty phone that he can barely use... what is he supposed to do? I could barely read the menu on my iphone 13 pro max...
Every time I go out to eat now I have buyers remorse. Except the Mexican restaurants...they feel pre-pandemic still.
I recently ate at a Brix. Before the pandemic it was solid, great quality and reasonable prices (wood fired everything, craft beers). If I didn't feel so bad for the old decrepit workers I would have cussed them all and ran out. Cost a fortune. Never going back.
In addition to the dumb menu thing, the food was terrible... to the point that the rice tasted like plastic. The prices were off the charts, but I'm used to that at this point. To add insult to the whole experience, tips were auto calculated on top of tax on the receipt (which was even noted on the receipt), just to boost it up higher.
So yea... lesson learned. I wonder if there are any statistics around post-covid return rates for restaurants. I bet it is terrible.
There's probably more takeout and you have to take into account the effect of inflation on dining as well. But the number of people dining out in the US is broadly back to pre-pandemic levels although people are eating out somewhat less frequently.
Interestingly, the place I went to last, was packed full in the late afternoon (~3pm), on a monday. The large bar seating was full.
My experience so far is that the quality of nearly every dish I've eaten has gone down significantly to the point that I keep trying new places and almost never repeating.
It could be that now that I cook more at home, I'm just not used to the amount of butter/salt that is being (ab)used in restaurants, as well as the quality of the ingredients (I mostly buy/eat "organic", whatever that is worth...).
In terms of the percentage people who dine in or get out at least sometimes numbers seem to be pretty much back to pre-pandemic levels.
That said, we do have laws in the US around accessibility for handicapped and disabled people and this could easily be added as an extension to that.
Had this been a place where there were no servers and you could only order and pay by phone (which also exists), then that brings up the question of exclusion of people who don't embrace technology or can't afford a phone. I've seen this one come up a lot on HN in the past too.
For old people you would need a plan though, like printed menus.
what is wrong with that?
It depends.
I like depositing checks by app rather than having to go to an ATM--much less a bank clerk.
And I'm generally OK with self-service checkout if I just have a few items and everything is barcoded.
But, yes, there are many cases where automation (e.g. chat bots) and self-service save the business money at the cost of making the consumer experience worse.
I spent years in Amazon Fulfillment Tech, a lot of in the Picking software space. It's all about tradeoffs, and shipping costs.
Amazon will spend far more money getting the item from the warehouse to your front door than they will on labour costs getting the item into the box, ready to ship to you. Shipping is expensive, labour is relatively cheap when each item is touched by maybe 2-4 people for a few seconds each.
Robotics are getting better, but the typical storage bin in an Amazon warehouse is optimized for density, not for easy picking. Humans do a great job of picking items despite the storage density. We fish it out, feel around, use our incredible vision powers. We're great at it! Robots suck at that sort of thing (today).
Amazon could build a warehouse that is optimized for robotic picking. But if you built it with today's robots, you would have a fraction of the item storage density. That means you'd either need many times more warehouses or you'd need to reduce your selection. When you reduce selection, that means a different warehouse is handling those other orders, and it's probably further away from the customer.
Further away from the customer => higher shipping costs.
I just read the article and does not talk about automation. Did I miss something?
Amazon has been working on robotics for years and has been incorporating it into their warehouses for a long time.
https://www.supplychaindive.com/news/warehouse-robotics-auto....
Not sure how much I agree with this, but a buddy of mine for years talked about how problematic it was for warehouse workers to fight for higher wages and more support because their jobs would likely be automated away in the near-ish future. His argument was that these jobs are not intended to be long-term career jobs and that by pretending they are, you run the risk of doing people a disservice as they stay longer and they become less employable over time.
They really don’t make significant use of robots.
No idea where this data is coming from. But there are a lot of factors at play and this is pretty high-level data on what could be going on. It doesn't account for changes in how the company operates, the number of warehouses that are spinning up, how packaging and shipping has evolved (moving between warehouses to consolidate multiple orders into single shipments, etc), the massive spike in orders during the pandemic...
But really, What's your goal here? It seems like you just want someone to say you are right instead of discuss? If you wanted to discuss, I'd expect that numbers thrown out would have some sort of backing behind them and displaying curiosity as to why things may be operating a certain way as opposed to already deciding how things work.
Tech workers in air-conditioned buildings are out of touch on how the underclass live and work. BUT I PICKED TOMATOES FOR A SUMMER JOB WHEN I WAS 15
I expect that in the (near?) future when AI has replaced software engineers, lawyers, writers, and bloggers that humans will still be needed to be warehouse workers, fruit pickers, plumbers, and surgeons.
The more I think about this, the more I believe you to be incorrect! Identifying the correct item in the bin is hard - things are in random arrangements and orientations. They may be buried under other stuff. Packaging continually changes. Things will be in the wrong bin. Things will be packaged by the manufacturer incorrectly! Things expire but you can't rely on expiration dates, which you'd have to be able to read and parse in the first place. Anything in the bin could be extremely fragile. Some things are both fragile and heavy.
The act of picking stuff up is hard. Grip too hard, you break it. Grip too soft, you drop it and break it or something else. Grip it the wrong way, you break it. Crush stuff around it rummaging, you break that. Some things must be pinched, others grasped, cradled or supported at multiple points. You break something, then what? You've contaminated who knows what around it roomba style - you can't send stuff out to customers thats covered in shards of glass or unknown liquids.
How do you identify when something is broken or defective in the first place? Packaging may be damaged while the item is OK, or the item may be broken while the packaging is intact. Humans can do this because they have a model of the world with interactive feedback. Things with broken glass may sound wrong or just feel wrong when moved.
The stakes are lower on average, but putting the wrong item in a cart will eventually kill someone. Consider allergies, medical supplies, the blind or illiterate. All the while, you have to do this for less than $15/hour? I can only see robots currently succeeding for hyper specific items - perhaps those that are low value and heavy for instance.