5 ms·

> On April 4, 2024, it was revealed that Amazon's "Just Walk Out" technology was supported by approximately 1,000 Indian workers who manually reviewed transacti

• theanonymousone 9mo ago
  Why did "outsourced workers get (relatively) much more expensive after"?
  • mjr00 9mo ago
    Great question. I'm not an economist so I have no idea why. The outsourcing rates I've all seen have gotten way higher in the past ~10 years though.
    • Insanity 9mo ago
      Beyond just the usual inflation? I'm not an economist either, but I also assume that as the country attracts more local talent for local companies, the competition for outsourcing becomes harder. (i.e, you now have to pay more than the local companies). All just speculation on my part though, I really have no clue either.
      • PaulHoule 9mo ago
        People from Bangalore were telling me it was getting crazy expensive to live there (by Indian standards) circa 2013.
  • foxyv 9mo ago
    Essentially the thinking went. If everyone is remote, why not hire remote workers from countries that are a lot cheaper. Suddenly you had a hard time finding contractors and FTEs from those countries because everyone was hiring them. At the same time it got really hard for entry level developers in the USA to find work. The supply/demand curve shifted and now those workers are becoming more expensive while domestic workers are becoming cheaper.
  • giraffe_lady 9mo ago
    India specifically is in the middle of a massive years-long labor movement that is changing the terms of work there and I believe shifting the degree of alignment with western corporate outsourcing though I'm not very informed about the details. Scale is beyond comprehension though, there were 250 million people on strike one day last summer. This is not ever really covered in western media or mentioned on HN for reasons that are surely not interesting or worth pondering at all.
    • givemeethekeys 9mo ago
      Americans can’t afford to strike like that.
      • esseph 9mo ago
        Can't afford not to.
      • linkregister 9mo ago
        You're most likely correct; I originally started writing this comment to refute your statement, but found that my assumptions appear to be wrong. Americans have the nearly the highest nominal and PPP income of OECD countries as of 2024, only behind Luxembourg, Iceland, and Switzerland [1]. India experiences substantially higher shelter and food insecurity and poverty rates than the United States. However, tech workers in Bangalore are paid an order of magnitude higher than prevailing local wages in other sectors, at around ₹2M (₹20 lakh) [2]. Median annual rents for 2BHK (2 bedroom) apartments appear to be around 1/10th of that figure at ₹3 lahk in desirable neighborhoods [3]. It appears to be reasonable for a technology worker to be able to perform a sustained strike. I have never personally traveled to Bangalore, though I have lived in places where cost of living is under a tenth of median American income. I invite correction by people with first hand knowledge about cost of living in Bangalore. 1. https://www.oecd.org/en/data/indicators/average-annual-wages.html https://www.oecd.org/en/data/indicators/average-annual-wages... 2. https://timesofindia.indiatimes.com/city/bengaluru/median-tech-pay-in-india-declines-sharply-us-hits-new-highs-report/articleshow/124856820.cms https://timesofindia.indiatimes.com/city/bengaluru/median-te... 3. https://www.birlaevara.org.in/best-areas-in-bangalore-for-rental-returns-and-passive-income.html https://www.birlaevara.org.in/best-areas-in-bangalore-for-re...
• thinkingtoilet 9mo ago
  Another case where AI = "actually Indians". It's funny how often this has happened.
  • Dylan16807 9mo ago
    Maybe. I'd really want to know what percent of items (not transactions) needed review. 1,000 people to oversee how much revenue? Theoretically if it was 99% computer and 1% human, that's enough to mess up the economics but it's not a bait and switch like some companies have done.
  • kkkqkqkqkqlqlql 9mo ago
    I remember this case the one who put "Actually Indians" in my mind. What other instances do you know? (Not to refute your point, of course, I am just curious)
    • Sateeshm 9mo ago
      Builder ai
• Cornbilly 9mo ago
  It's great that they faced essentially no consequences for this. A sure sign that we have a functional and sane market.
  • dyauspitr 9mo ago
    It’s autonomous 80% of the time. That’s significant. Put another way, they only had to hire 1000 people instead of 5000.
    • mrguyorama 9mo ago
      It only takes 1 employee to staff 20 self checkouts for comparison. For a full fat grocery store. With zero change or adjustment to the rest of the grocery store. And customers weirdly like self checkouts even when they are a dramatically worse outcome (compared to the highish bar of well trained cashiers)
      • dyauspitr 9mo ago
        What exactly is the point of a well trained cashier, what service do they provide. I guess I appreciate what the bagger does and the cashier knows the codes for the loose vegetables but those are minor benefits in my opinion.
      • octoberfranklin 9mo ago
        We like self-checkout because there's hardly ever a line. An idle self-checkout machine costs the store almost nothing. An idle cashier costs the store wages. So the stores will always skimp on cashiers, leading to lines, wasting my time.
  • colinplamondon 9mo ago
    Why would they face consequences? Every store has video surveillance that can be reviewed. They trusted their tech enough to accept the false-positive rate, then worked to determine / validate their false positive rate with manual review, and iterate their models with the data. From a consumer perspective the point is that you can "just walk out". They delivered that.
• ed_mercer 9mo ago
  This was proven to be false on the WAN show. Only 20% of transactions were low confidence and handled by mechanical turk. https://m.youtube.com/watch?v=433kipkEERY&t=8479s https://m.youtube.com/watch?v=433kipkEERY&t=8479s
  • larrik 9mo ago
    20% seems like a "significant portion" to me
    • Breza 9mo ago
      For sure! Twenty percent moves you from "Game changing tech" to "Slightly improved self checkout."
  • mjr00 9mo ago
    20% is an incredibly high number though, if a store has 400 people/hour that means you're manually reviewing 80 transactions per hour, over one transaction per minute. That's multiple human employees.
    • iLoveOncall 9mo ago
      One transaction per minute is nothing at all when the transaction can be as simple as "did the person put that back on the shelf" with a 5 seconds clip.
      • freejazz 9mo ago
        If it was clear from just a 5 second clip it probably wouldn't have needed to be reviewed
        • iLoveOncall 9mo ago
          Hum, yes it does? It's not because it's not a complex action that it's necessarily supported by the models. It's not hard to imagine edge scenarios for which the models aren't trained, like a customer dropping an item, or putting an item back in a random shelf instead of the one it's intended for, or someone picking up that previously randomly placed item, etc.
• adamsb6 9mo ago
  People don’t know what the H is in RLHF.
• davidst 9mo ago
  I left the following comment some months ago, duplicating it here: [Disclaimer: Former Amazon employee and not involved with Go since 2016.] I worked on the first iteration of Amazon Go in 2015/16 and can provide some context on the human oversight aspects. The system incorporated human review in two primary capacities: 1. Low-confidence event resolution: A subset of customer interactions resulted in low-confidence classifications that were routed to human reviewers for verification. These events typically involved edge cases that were challenging for the automated systems to resolve definitively. The proportion of these events was expected to decrease over time as the models improved. This was my experience during my time with Go. 2. Training data generation: Human annotators played a significant role in labeling interactions for model training-- particularly when introducing new store fixtures or customer behaviors. For instance, when new equipment like coffee machines were added, the system would initially flag all related interactions for human annotation to build training datasets for those specific use cases. Of course, that results in a surge of humans needed for annotation while the data is collected. Scaling from smaller grab-and-go formats to larger retail environments (Fresh, Whole Foods) would require expanded annotation efforts due to the increased complexity and variety of customer interactions in those settings. This approach represents a fairly standard machine learning deployment pattern where human oversight serves both quality assurance and continuous improvement. The news story is entertaining but it implies there was no working tech behind Amazon Go which just isn't true.
  • londons_explore 9mo ago
    As soon as you get to ~99% accuracy, you probably don't need to go further. If the customer is accidentally billed for an orange instead of a tangerine 1% of the time, the consumer probably won't notice or care, and as long as the errors aren't biased in favour of the shop, regulators and the taxman probably won't care either. With that in mind, I suspect Amazon Go wasn't profitable due to poor execution not an inherently bad idea.
    • Slartie 9mo ago
      Actually, discount grocers operate on razor-thin margins of 2-4%. If your inaccuracy is geared to the benefit of your customer (because otherwise you'll be out of business due to the regulatory bodies) and thus removes just one percent of that, you suddenly lose a quarter to half of your earnings! And that goes ON TOP of the additional cost incurred with all that computer vision tech. In addition to that, you'll have the problem of inventory differences, which is often cited as being an even bigger problem with store theft than the loss of valued product. If the inventory numbers on your books differ too much from the inventory actually on the shelves, all your replenishment processes will suffer, eventually causing out of stock situations and thus loss of revenue. You may be able to eventually counter that by estimating losses to billing inaccuracies, but that's another complexity that's not going to be free to tackle, so the 1% inaccuracy is going to cost you money on the inventory difference front, no matter what.
• gorgoiler 9mo ago
  I’m skeptical of this scoop. It’s reasonable to expect a system like Amazon’s to use human feedback in training, and to quote the article linked on Wikipedia: > Amazon said that the India-based team only assisted in training the model [and validating] a small minority of shopping visits.
  • hereonout2 9mo ago
    I went to Lidl UKs first walk out shop a few weeks ago. You get the bill and receipts about 40 minutes after you've left. It certainly felt like it could have been sent off to a lower paid country for a human to tot up. Also consider you're in the store for what, 10 mins - that's a lot of video processing presumably using state of the art CV models. It's quite possibly cheaper to pay a human than rent the H100 to do it.
• andoando 9mo ago
  I wonder if they were doing the same thing for palm recognition