Live data from Hacker News

Amazon hid its safety crisis

revealnews.org

71–80 of 143 posts

Re: Amazon hid its safety crisis

#71
post #15

Earlier quoted context omitted.

Well, until it's fully automated, human safety is still important. I think the problem we have is that destroying a human body with mindless repetitive tasks is often still "cheaper" than engineering an automated solutions. It is simply cheaper to feed off desperation than to value human lives, and there is the engrained notion that humanity cannot be productive without putting humans into positions of desperation. T…

When I was a graduate student, the lab next door was working on robotic arms and picking challenges, their PhD grad was then hired by... Amazon. I don't think it's cheaper, it's just that a solution has yet to exist. Once it start working, it will go out at breakneck pace.

> I don't think it's cheaper, it's just that a solution has yet to exist. Once it start working, it will go out at breakneck pace.

I don't think this is a single problem. There are several dimensions to it, and you can trivialize some at the expense of others. Stop storing things compactly in bins and you can make it a lot easier to grab those things with a robotic arm. Give everything very robust packaging and you no longer have to worry about appropriate grip strength.

Re: Amazon hid its safety crisis

#72

> Jeff Bezos’ 14 leadership principles are famous inside and outside Amazon for vividly articulating what is expected of the company’s leaders. The first is customer obsession. “Leaders start with the customer and work backwards,” it reads. And yet Amazon has no real interest in solving its counterfeit problem, the same way it has no real interest in solving its safety problem. I feel like it's all PR and bullshit at…

No interest in solving the problem? I work at Amazon, we are hiring all the top ML people to work on this problem. Legions of ML PhDs. Its not easy and you are just making stuff up

How hard can it be? Keeping inventory from different sources separate should go a long way to fixing things. Right now, as a customer I can't rely on the trustworthiness of the seller because they may not be the one who supplied the product I ultimately get.

Re: Amazon hid its safety crisis

#73

> Jeff Bezos’ 14 leadership principles are famous inside and outside Amazon for vividly articulating what is expected of the company’s leaders. The first is customer obsession. “Leaders start with the customer and work backwards,” it reads. And yet Amazon has no real interest in solving its counterfeit problem, the same way it has no real interest in solving its safety problem. I feel like it's all PR and bullshit at…

I understand the sentiment but for the records: I don't think employees are the customers of Amazon. I definitely don't think Amazon sees it that way.

What you said makes more sense if you replace both the words with the word 'humans'.

Re: Amazon hid its safety crisis

#74

Earlier quoted context omitted.

This is why I see capitalism eventually failing. On one hand, I think it's great that these mind-numbingly boring jobs can be automated away. On the other hand, the end result is that more profits go to Amazon, while many more people have fewer opportunities to participate in the economy. For the old predictable arguments of "Well, these people can do something else, do you expect buggy whip makers to still be around…

That’s why we have a government, to adapt society to an ever changing world. You can postulate as to _if_ government will step in an enact the requisite policies (UBI, Retraining, ect) but it doesn’t take a big imagination to see a world post-automation. There’s even a whole TV series set in such a world: Star Trek.

Sure would be cool if we could skip the whole dystopian bit, and the WWIII part, and the genetically engineered people rebellion.

Re: Amazon hid its safety crisis

#75

Earlier quoted context omitted.

This is why I see capitalism eventually failing. On one hand, I think it's great that these mind-numbingly boring jobs can be automated away. On the other hand, the end result is that more profits go to Amazon, while many more people have fewer opportunities to participate in the economy. For the old predictable arguments of "Well, these people can do something else, do you expect buggy whip makers to still be around…

The response to this is usually, "Well, then people should be retrained." It's kind of an elitist, SV bubble trope at this point. There are millions of people who simply cannot be retrained. Yes, pick-and-pack is mind-numbingly boring for many people. But for others, it's good. Or even challenging. I've worked on an assembly line three times in my life. I found it tedious and pointless, and to keep my mind busy, I mo…

> The solution is to automate an appropriate amount of factory jobs, but also keep an appropriate amount of low-skill jobs so that these people can be part of society.

Yeah but, isn't the fact that those are going away the whole problem? If there already was an "appropriate amount of low-skill jobs" we wouldn't be talking about this. Do you want the government to mandate certain companies have to keep X number of certain jobs around for low-skilled workers? And what do you do about jobs that simply aren't needed anymore (not automated away, just gone due to something new)?

I'm not saying it's a terrible idea, but it seems odd to me that you're complaining about UBI but your solution to the problem is likely just UBI with extra steps - is it more 'demeaning' to just give people money for nothing, or to force them to do a literally unnecessary job to get it?

It's also not like there isn't anything for people to do if they're receiving UBI in place of their (now gone) job, Ex. they could volunteer at a non-profit, work on hobbies, work for a local business (that maybe couldn't afford to hire them at a full salary), etc. And all those things are arguably more useful usages of their time than making them do an unnecessary job.

Still, fundamentally I agree with most of your criticisms, there's definitely a subset of the population that really can't be retrained from what they currently do, but recognizing the problem and solving the problem are two different things. And numbers here matter as well - the number of people losing their jobs to automation is a different number from those that can't reasonably be expected to be retrained somewhere else. If that second group is small enough, just giving them UBI/SS/etc. and telling them to retire is honestly not that bad of an option.

Re: Amazon hid its safety crisis

#76

> Jeff Bezos’ 14 leadership principles are famous inside and outside Amazon for vividly articulating what is expected of the company’s leaders. The first is customer obsession. “Leaders start with the customer and work backwards,” it reads. And yet Amazon has no real interest in solving its counterfeit problem, the same way it has no real interest in solving its safety problem. I feel like it's all PR and bullshit at…

No interest in solving the problem? I work at Amazon, we are hiring all the top ML people to work on this problem. Legions of ML PhDs. Its not easy and you are just making stuff up

> No interest in solving the problem? I work at Amazon, we are hiring all the top ML people to work on this problem. Legions of ML PhDs. Its not easy and you are just making stuff up

I think you may be a touch defensive here. Take a look at what makes the the problem hard -- it's the incentives, not the lack of PhDs. The problem is that Amazon makes money off the counterfeits until caught, and when caught the penalty is low enough that it's not worth it to the GM to prevent the problem.

If you truly have an interest in solving the problem, take up that incentive problem with the GMs in charge and see how far you get.

Re: Amazon hid its safety crisis

#77

Earlier quoted context omitted.

Some guy below you: > I work at Amazon, we are hiring all the top ML people to work on this problem. Legions of ML PhDs. It doesn't take legions of PhDs in machine learning to fix a basic counterfeit problem on a glorified online shopping store.

Amazon doesn't even let their customers choose the correct reason for returns when they receive counterfeit items. Instead of being able to choose " This item is a counterfeit " or " I believe this item is a counterfeit ", the customer must choose between reasons for their return that aren't entirely accurate, like " Inaccurate website description " or " Item defective or doesn't work ". You'd think that a company th…

It's like politicians preventing studies on racism in police (a current hot topic in Germany): As long as there is no data, no one can demand action (i.e. getting rid of racist cops) or subpoena it (i.e. a class action lawsuit against Amazon).

Amazon has a vested massive interest in not provably knowing numbers about counterfeit rates!

Re: Amazon hid its safety crisis

#78
post #61

Earlier quoted context omitted.

THe Kiva video is not state-of-the-art I think. Check out this automated grocery warehouse, something right out of Factorio: https://www.youtube.com/watch?v=4DKrcpa8Z_E Humans would just get in the way and get mangled.

That's a rendered video of the Ocado warehouse system, which I think is a ripoff of the Swisslog Autostore system. There was a lawsuit. Then Ocado hired Swisslog... Btw: This Autostore system is becoming the standard automated warehouse system for Swedish e-retailers, as they prepare for Amazon entering the Swedish market in a month or two. I think there are at least like 10 very large Autostore installations in Swed…

I'm amazed that Swisslog system works reliably. All that precision track and rack. All those moving parts. There have been big automated storage and retrieval systems for decades, but they're often a maintenance headache.

The Kiva system just needs is a flat floor. The robots are mechanically simple and totally interchangeable. On-site maintenance is limited to replacing batteries and wheels; for anything else it goes back to a repair center.

Re: Amazon hid its safety crisis

#79
At the risk of defending a faceless megacorporation, I'm curious what injuries per package look like at Amazon vs "Industry Average".

According to the article, productivity expectations per worker doubled with robotics, and Amazon has about double the injury rate per worker, so my super rough back of the envelope calculation says they should fare similarly on that metric.

If the industry alternative takes twice as many workers to do the same job, ultimately leading to approximately the same number of injured workers, has Amazon really created a safety crisis?

Re: Amazon hid its safety crisis

#80
Speaking of Amazon trying to downplay its wrongs, I recall there was a post here that counted Amazon PR team at 400-odd people. But I could not find it. Are they so effective or have I Mandela'd myself and this did not happen?
Post reply on HN