Welcome to Incremental Social! Learn more about this project here!
Check out lemmyverse to find more communities to join from here!

Techranger ,

It would be fun to have a Bollywood style flashmob show up at one of these stores.

Rooskie91 ,

Hey! Guess how they remove all of the content they DON'T want to train AI's on?

A disturbing portion of the internet's scams are held up by schemes like this.

Lev_Astov ,
@Lev_Astov@lemmy.world avatar

What ever happened to doing this with UHF RFID? Getting the cost of the individual chips down was always just a matter of scaling production.

londos ,

They weren't just tracking what you bought. They wanted to track what you looked at and for how long, to learn what packaging worked best.

just_change_it ,

Jesus christ these headlines mislead everything.

They were using machine learning to try and figure out what people were buying. Machine learning has lots of errors until you train it. The "hundreds of workers" were training it by telling it what each thing was. E.g. it was creating training data for it to learn from.

The goal was to train ML enough so that humans were rarely necessary, obviously.

MajorHavoc ,

The goal was to train ML enough so that humans were rarely necessary, obviously.

Yes, that's the goal.

There's a long rich history of AI like outcomes being mimiced by just hiding the human who does the work. That's actually the source of the name of Amazon's own "Mechanical Turk" service.

Not being actively watched by an army of underpaid workers is effectively still on the "someday...maybe" feature list for this thing, unless Amazon (famous for making delivery workers pee in soda bottles, and allowing warehouse workers to get heat stroke) somehow provides credible proof that they've actually grown past that.

I, as someone with substantial professional ML experience, won't take Amazon at their word, when they claim the ML has alleviated the need for the army of workers watching cameras. That's bullshit marketing promise, until proven otherwise. Particularly coming from Amazon.

Moving away from the people watching to using pure AI is well within the realm of possibility.

But good AI maintainers cost more per hour to pay than the entire army of mechanical turk "trainers". So I am skeptical of any claim that Amazon, in particular, has done the right thing here.

So it's very fair to assume you're being watched in one of those stores, until real credible evidence is provided that you're not.

baru ,

Jesus christ these headlines mislead everything.

One article included how often employees needed to look at the cameras. That was the case in something like 80% of the times people went in to shop.

The goal was to train ML enough so that humans were rarely necessary, obviously.

The headline is pretty accurate. That might have been the goal, but they didn't come close. And now they are closing down those stores.

Seems that they utterly failed in the goal.

Machine learning has lots of errors until you train it.

These stores were open for a pretty long time. It's not a given that it's just a matter of training.

RealFknNito ,
@RealFknNito@lemmy.world avatar

The people who hate AI always seem to have no fucking idea how it actually works and it's frustrating.

People were required to teach the AI how to do it's job. A 'new employee' is going to make frequent mistakes during training. Yes - it took a long time to train a program to identify a person, match them to their ID, identify a product, match it to the UPC, then make absolutely 110% sure that item remained on the person when they left with no mistakes. For it to be flawless even when you shove it into your backpack.

Should the people training it have been paid more for their temporary position? Sure. Should Amazon have been transparent about how they were teaching the AI? Sure. They still did not rely on Indian workers for their stores, they relied on people to teach the AI, like doing a captcha, that the store relied on. Any human being would have done the trick, India just allows its people to be exploited the most apparently. Headlines are meant to make you click, not give you accurate information.

Experiments in technology don't always work. This was a bold plan that they gave years to which would have been a really cool thing to have. Just grabbing your shit and leaving? That's like EZPass for retail. There was definitely money there, they just couldn't get to it in time.

steeznson OP ,

That's their excuse but it is convenient for them that in order to train the AI the workers need to follow the exact same steps as what an AI would be doing if it was sufficiently trained. We can't say as outsiders to what extent the actual work is assisted by AI. Seems likely that it is largely a manual process.

Railcar8095 ,

I understand the spirit, but that's how it goes. You have somebody doing the work, as you want the ML to do it, and then feed the data. It's the same when they get oncology scans that have been diagnosed by well paid doctors, somebody who knows does and the machine tries to replicate.

What very likely happened is that the failure rate platoed much higher than they expected, and all this time the goal was to lower it. Remember, it's cheaper to have 0 people in India than 1, specially with AWS in mind.

Moreover, even if the accuracy was incredibly high, they would still need people reviewing. You have to review random events to ensure the model keeps performing well and to evaluate the ones with low confidence or suspicious.

Zron ,

My goal is to build a fusion reactor.

I will hire Indian call center workers to add fuel to my diesel generator until the fusion is up and running.

This plan makes sense to certain people on the internet.

AnUnusualRelic ,
@AnUnusualRelic@lemmy.world avatar

You could probably get lots of funding with this. I say go for it.

BakerBagel ,

Except they still had thousands if employees in India watching the surveillance tapes to see what people bought and charged them for it

Amazon can claim this was a stop gap all they want, but the truth is that the technology behind the core concept isn't there and they just pretended it worked so the project head wouldn't have to explain why they are behind schedule and over budget. It's the same as with their drone delivery service 10 years ago. All smoke and mirrors to make moron tech bros cream themselves

echodot ,

You need training data though I don't understand what the problem is. Hell it doesn't even matter if they never actually make the technology work that'll be their problem. They can't lie and tell people it works if it doesn't but as far as I'm aware they're not actually doing that.

I don't like amazing very much but they do enough crappy things for you to actually get upset about so it just seems odd that you would pick this hill to die on.

BakerBagel ,

They gave up collecting that data though, that's why they are shutting down the department. All they managed to do was outsource grocery store cashiers to India, which seems like an exceptional shitty thing to me

TurtleJoe ,
@TurtleJoe@lemmy.world avatar

They were using machine learning to try and figure out what people were buying. Machine learning has lots of errors until you train it.

Machine Learning, no matter how well trained or advanced, is just doing a make-em-up.

Besides that, in this case the experiment has been going on for years and humans were still doing like 70% of the work. It was a failure, that's why Amazon shut it down

Gigan ,
@Gigan@lemmy.world avatar

They found a way to outsource grocery store employees. great

postmateDumbass ,

If we are in an era of economic wars, can we get laws to charge companies with treason?

Krauerking ,

Yeah this has been my biggest takeaway actually from this story circulating.

The rich and business owners have now found a way to export even basic service jobs to cheaper countries to mostly save themselves the money and reap the profits.

I'm a pretty far leftist but holy moly we really need to stop selling every job to cheap laborers because it's only enriching those who already have enough and seems to do little good for anyone else.

towerful ,

I dont think "not wanting to outsource to the cheapest labour pool" is a hallmark of the left or the right wing.
I think "not wanting to outsource" is both progressive and anti-capitalist.

Doing things as cheap as possible, no matter who suffers, is absolutely a capitalist thing, and is why government regulation and government oversight is so important... As long as the government isnt in on it, of course (thats just a corrupt government)

FlashMobOfOne ,
@FlashMobOfOne@lemmy.world avatar

I dont think "not wanting to outsource to the cheapest labour pool" is a hallmark of the left or the right wing.

Exactly. At least here in the US, both ruling parties are 100% capitalist and exist to enrich the wealthy to an even more obscene degree than they already do.

It's why when they ask for political donations I put that money in stocks instead.

SparrowRanjitScaur ,

That's not the point at all. They wanted to automate this with computers and it failed. They had to rely on workers in India to do the analysis. That's why they're canceling this service.

dustyData ,

But AI is intelligent and sentient, just throw more server farms at the problem, right?

/s

SparrowRanjitScaur ,

I know you're trying to be funny, but while AI does have many uses, I imagine the issue for Amazon was that much like self driving there were just too many edge cases to try to solve here.

Telodzrum ,

All "AI" relies on thousands and thousands of low-paid, overworked humans in the Global South staring at screens so snake-oil salesmen in SV and Redmond can claim they're about to revolutionize the world. It's all lies.

GenderNeutralBro ,

All “AI”

Not even close to true.

Telodzrum ,

If you unhinge your jaw a little further I'm sure you can fit that VC cock further back in your throat.

GenderNeutralBro ,

I think you are confused about what "AI" means. You are referring to a very small subset if AI.

feedum_sneedson ,

That's got nothing to do with how wide my mouth is open.

echodot ,

Actually most AI relies on high quality data acquired from all sorts of sources including academic research papers, GitHub code repositories, books and journals, newspapers, and technical manuals.

Krauerking ,

Uh-huh... And yet reddits biggest source of cash recently was selling the comments section for training AI data... Sure. "High Quality"

Only if you recognize most the comments are written by people who are high.

echodot ,

Reddit is run by an idiot. Don't worry about what reddit's doing they're not making money.

Telodzrum ,

Manually parsed and corrected by near-slave labor.

Enjoy your funny ChatGPT output though, you monster.

echodot ,

What the actual hell is that comment? I'm sorry are you suggesting some kind of evil cabal or something?

Might want to make some more sense

stevedidwhat_infosec ,

Big fuckin shocker

You mean to tell me big businesses are exploiting low wage countries?

bangupjobasusual ,

I’m pretty sure the first chess program was just a grand master playing remotely

steeznson OP ,

Yes, that's where amazon got the name of their service from

brsrklf ,

Well, that one was not so much remote and more like hidden in the silly fake automaton thing.

ViscloReader ,

The type of shit you'd see on a movie

_sideffect ,

"Hey Vitesh, this John guy just stole two chickens"

Granixo ,
@Granixo@feddit.cl avatar

I knew it!

steeznson OP ,

Yeah it was always suspect when this "shopping experience" came from the company behind Mechanical Turk.

BobDolesBBallHandle ,

Was about to make the same comment 😉. Will share the wiki instead.

https://en.wikipedia.org/wiki/Mechanical_Turk?wprov=sfti1#

snooggums ,
@snooggums@midwest.social avatar

I thought it was China.

steeznson OP ,

I'm not exactly sure how they would have set this up given that their usual solution of Mechanical Turk does not pay their workers in cash anywhere outside of the US.

In some rural parts of the states Mechanical Turk is the largest employer but workers in other countries can only get paid in Amazon vouchers.

Somehow there are still a lot of Indian people working for Mechanical Turk though. It's not clear if they are exchanging the vouchers or are stuck in a hellish walled garden where their wages can only be spent with their employer.

snooggums ,
@snooggums@midwest.social avatar

a hellish walled garden where their wages can only be spent with their employer.

There are also a bunch of reports of businesses in India where people with loan debts being forced to work in hellish conditions and not allowed to leave by armed guards, basically indentured servitude/slave labor.

feedum_sneedson ,

Isn't that like, company scrip. I thought that was meant to be illegal.

BakerBagel ,

It is, but Uncle Sam can't do anything about how you "pay" your "workforce" in other countries.

I mean, they actually can, but then corporate lobbying groups will be sad, so Congress wont do anything about it.

autotldr Bot ,

This is the best summary I could come up with:

The company touted the technology, which allowed customers to bypass traditional checkouts, as an achievement powered entirely by computer vision.

An Amazon spokesperson disputed that claim in a statement to Business Insider, saying that the team in India mostly helps train the model that the company used for Just Walk Out.

"Associates may also validate a small minority of shopping visits where our computer vision technology cannot determine with complete confidence an individual's purchases," the spokesperson said.

While customers used Just Walk Out at Amazon Fresh stores, "they also wanted the ability to easily find nearby products and deals, view their receipt as they shop, and know how much money they saved while shopping throughout the store" — all options that the company's Dash Cart provides, the company spokesperson said of the change.

The technology allowed customers to enter a store by identifying themselves with their Amazon account.

Startups have also created their own versions of Just Walk Out and tested them at retailers including Aldi and Dollar General.

The original article contains 432 words, the summary contains 169 words. Saved 61%. I'm a bot and I'm open source!

JohnnyCanuck ,
@JohnnyCanuck@lemmy.ca avatar

Wait, how does this technology work!?

(other than being really bad at what it does... )

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • technology@lemmy.world
  • incremental_games
  • meta
  • All magazines
    •