Amazon competitor got our belt listing taken down by saying it contains drugs
travelhead.medium.com
travelhead.medium.com
Of course none of my competitors using the exact same data set had any such problems.
I tried for YEARS to appeal it. There are simply no humans working at Google and nobody reads your emails.
Edit: Actually, I did get a response a couple times but it was obviously automated. They just said to remove the ads from the pages where such words are displayed. So I added a simple rule and a column in the database to hide ads for those keywords. That just triggered the bot to move down the list of their "obscene" language. Next it was the names of various sexual positions, acts and fetishes (Japanese does have a very rich vocabulary in that topic), then manga slang, even silly sounding onomatopoeias that when explained in plain English are "vulgar", etc.. It seems once your website is flagged there is simply no way get clean.
:(
I realize it's complicated, but there are things that make me worry the future of the internet might not be so great.
Imagine the web in the 90s, some people saw promise while others asked themselves what it actually is. Fast forward 5/10/20 years and what web3 is will be a lot more clear as we'll have mainstream web3 applications deployed and used then.
While blockchain based technologies are often cited as an example of Web3, and some of the worst hype offenders are pushing this. It's not exclusively blockchain.
In some ways Web3 is a return to earlier internet where there were open protocols rather than everything being controlled by companies. Redecentralize[0] has a great set of founder interviews talking about some of the projects and challenges.
To be clear it's not necessarily anti-company, just a recognition that a handful of companies controlling everything is problematic. Coupled with a recognition that running at the scale of the Big 5 is prohibitive (e.g. server costs, preventing malicious actors, people to handle problems).
The UX around these things is hard, though getting better. Mastodon[1] is a reasonable alternative to twitter now (doubly so if you cross-post).
Gitcoin[2] is a good example of a Web3 app (all the core functions are smart contracts on the Ethereum blockchain).
There are plenty of example of Web3 apps(dapps) on Webby[3] and DApp.com[4] (though the latter you have to wade through all the Decentralized Finance stuff).
[0]: https://redecentralize.org/interviews/ [1]: https://mastodon.social [2]: https://gitcoin.co/ [3]: https://heywebby.app/webby [4]: https://www.dapp.com/topics
However, it seems like blockchain-based technologies are to be an integral part of web3, or are there alternatives to it? Is there a way to talk about the concept without mentioning blockchain?
That's why efforts like redecentralize are so important. It's more talking about objectives like privacy, censorship resistance, etc.
There are a subset of people that go, let's build with "the blockchain" for the hype and another subset that recognizes that blockchains are one of a number of tools in the arsenal of solving a problem.
Efforts like Matrix, SSB, OpenBazaar, and IPFS are all 'Web3' technologies without being on the blockchain.
If you do need to talk blockchains, two terms that likely aren't so loaded are 'smart contracts' and 'dapps' (distributed apps).
Blockchain is a very heavy technology that solves the Consensus Problem. That's amazing, but almost no real world problems except cryptocurrency has a Consensus Problem.
For almost everything else, there are much simpler, easier-to-understand and _far_ computationally lighter cryptographic solutions that aren't blockchain.
He went on and on about double spending and consensus. i asked him why anybody, even his “ico investors” would care about any of that just to play a (slow) game.
He went back into double spending and byzantium generals and using Brave for tokenized payments.
This pattern repeated with every single dapp I came across.
* It's possible to gamble even in jurisdictions that make it illegal.
* The rules are encoded and visible, the house might always win, but at least they're not cheating.
The former point almost certainly makes small disposable servers in a variety of jurisdictions a requirement. There might be many millions of crypto users around the world, but there are very few running their own nodes, and fewer still actively mining. Block generation is practically as monopolised as the banking system it supposedly replaces.
To the latter, I worked a cryptocurrency casino and the provably fair algorithms were implemented using the core Node.js crypto libraries and completely independent from all the blockchain nodes. The sheer frequency of by users completely rules out any chance of leveraging a smart contract or block data as a base for the RNG.
Turns out being able to pirate MP3s is a poor consolation prize when you can have your startup permabanned from AdSense or Amazon for no good reason, and there's nothing you can do about it.
Wrong goal on the wrong sports field.
What do you think the free software movement was all about, if it wasn't about avoiding Kafka-esque monopolistic bureaucracies? Read Stallman's "The Right To Read", from 1997.
It's that Data Scientists don't realize most of their models still simply fit a conditional expectation - and given all their power, it's going to be us, not Amazon or Google, who has to adapt to the distribution as not to repeatedly get hit as some sort of outlier.
It's a dystopia where AI works because of us trying to conform to it, because otherwise we are out of luck. At some point, we self select into Amazon-humans, Google-humans etc.
But all technology is like that. Cars work, because we conformed our environment to it. Email works, because we conform to checking our mailboxes regularly. Telephones work, because we pick them up when they ring.
If you are interested in knowing more, the whole work of the french philosopher Jacques Ellul is based around this idea.
AI is not going to be any different in this respect. It is going to be us meeting it halfway. And that is going to be dystopic in the same way as it has been with other types of technology.
You also touch on the Amish approach to new technology. Each new thing is assessed for impact on life/tradition/etc and adoption is limited accordingly. For example, phones might be considered useful enough to have one per town, but distracting enough to be kept out of individual homes. Or, a computer might be allowed in the back-office of a wood shop for purposes for managing online sales, but nowhere else in the community.
With "AI" (machine learning / neural nets), we're letting opaque boxes arbitrarily control and ban accounts at will. It's like forcing people to use cars that randomly decide that it won't drive that day for no good reason.
"AI is not going to be any different in this respect"
The other examples all include accountable humans, "AI" here does not. We're already seeing issues of machine learning applying sexist biases for example because of datasets and poor training. What are you going to do, hold google, or any other faceless corporation, accountable for its bad translations? Good luck.
It's only going to get worse in the future as other stratified aspects of society get cemented by "AI".
I think one might underestimate the freedom that have been let go when other technologies came in, even though those technologies feel established already. You can choose not to pick up on the unknown number when the phone rings, but the moment of connection with a real human being you might have had at that point was interrupted by the phone ringing.
Society expects us to be reachable at any time, but that came with trading in the freedom to not be arbitrarily disrupted by default.
The point I am trying to make is this: all technology requires and has required humans to adapt to it, and consequently handing in a bit of freedom. I want to point out that AI is no different in the fact that it does so, even though like every other piece of technology it will do so in a unique way.
Is that dystopian? Perhaps. Many philosophers on technology have thought so, and many have disagreed.
> The other examples all include accountable humans
But when those faceless corporations were being sexist because of bureaucratic reasons in the sixties, how were they being held accountable? Did anything really change?
Widely adopted technology is something that lies outside of the humans that make it up, like an ant colony is different from the ants it is made off. You cannot hold any individual ant accountable to the way the hive is organised. Technology is an entity that lies outside of those ants, and that has its own desires and goals.
That's one worst case scenario. I've been thinking once humans start deselecting Google (no more gmail, youtube red, etc) then there's more room to also start blocking google's ad network. I know Google's dark pattern to force compliance is to pretend they don't know you and put you on a captcha treadmill, but that too will crumble if their adsense network yields less and less value as users and website operators move on from being plugged into the googleverse.
At least today you can operate outside the googleverse but you're punished for it. We should be holding on to that and trying to expand it rather than end up in the future you described where we can do nothing but comply.
The solution isn't easy, but it's simple: stop using Amazon (specially AWS), Google (specially AdSense), etc. Once they lose half their annual revenue, they'll start taking things seriously.
> Stop eating calories and you'll lose weight
As America approaches 50% obesity
> Stop using AWS, Microsoft Office, AdSense, and Instagram (If we want to talk about domineering monopolies)
Sure. Let me count the ways you've inconvenienced me.
A better approach would be to provide a viable alternative to change the underlying systems that put us in this state. That's never simple though.
So why even suggest it?
Strong regulation is the only solution. The Magic Invisible Hand just is not working.
Even bigger newspapers sometimes don't have ads on the more morbid type of news pages.
It may be questionable to have your stuff advertised on a page about the history of Germany...
But lets be honest... that's not why these things are being taken down. They aren't being removed because advertisers are complaining. Please show me where anything of that sort is mentioned - either in the OP of this thread or specifically in the article about "buckle-less belts"...
In the example of this thread... they are being removed because someone (presumably a competitor) reported their app for SUPPORTING "rape and pedos". A dictionary with a definition of rape getting removed for "supporting" rape by defining it - while competing dictionaries aren't removed for also defining the same words? Why aren't they getting removed for having a definition of rape that advertisers don't want to be listed on?
Please... show me where I'm wrong that the dictionary application was removed for 'advertiser complaints'... or the original article has its stuff removed for advertisers complaining.
In the long run, it has to make ad financed content less viable. We'll end up with advertiser friendly content, with no consumers.
Why am I seeing the words "Hugging" and "Cuddled" in summaries? The TV Calendar is monetised using Google Adsense. As part of Googles drive to be more advertiser friendly (you may know this as the Adpocalyse - it didn't just affect Youtubers) their system is now telling me the Calendar has SHOCKING CONTENT on it, and that ads will be removed from such pages.
[0] https://www.pogdesign.co.uk/cat/frequently-asked-questions.p...
* - at least not bigger issue than seeing any ads at all, but those would run ad-blocker anyway.
He is specifically taking about pages with Ads on them. Corporations don't want to be associated with said terms, so they wouldn't want to be advertised in that context.
The issue here is that Google, Amazon or $bigcorp allows its "blocking" algorithm to be misused, to be weaponized by competitors and that it allows no possibility of human review, that you can't get any human via phone or email to review a completely arbitrary (and often enough plain wrong) decision a random algo has committed (through human instigation or on its own rampage, as it happens often enough). That nobody has to account for and take responsibility for algorithmic decisions ("Sorry, computer says no. Good bye).
I have become so frustrated with this issue (we seem to discuss it every other day on HN) that I've come to think that you should not be allowed to offer services using automation without a clear and working process of escalation to a human in case of trouble. If we allow this behavior to continue and to spread (and look, I get it, doing business this way is very attractive and scales beautifully), things will only get worse.
We'll get to the point where it's "sorry, your car was deactivated because $passenger reported that you used a bad word in his presence. Do not contact us again". Maybe this will wake people up and show them that this way of offering services is broken.
Or "sorry, our algorithm decided you get the death penalty based on 100 reports of people that you did $badthing. chops head off Sorry, don't contact us again." Too much hyperbole? We'll see. Algorithmic determination of prison sentences already exists.
we only know what the opener said about his story, so i dont feel comfortable saying wherever what happened to him was dystopian or not.
That puritanical, prudish attitude has done so much damage already, we should stop accepting it as "Americans gonna American".
Is the problem that Google caries with it the puritan American culture, or that too many people which might not share that culture rely on Google?
I'm certain most Americans agree with me here which makes me assume that this specific problem is not one of culture, but probably scale and an absence of responsiblity at Google.
The more general question is interesting though, because it could go several ways. For example:
You could argue that people should anticipate that the platforms they rely on are not under their control (and should maybe act on that).
Or one could argue that the platforms should anticipate the diversity of cultural standards they are catering to by easing their moral rigidity. (For example through a more diverse/decentral company structure, etc.)
Here in Europe, some approach a somewhat similar question with some form of data nationalism, for better or worse. It plays into the same realization that there is an unresolved cultural difference between global platforms and local standards and intends to politically support local initiatives, corporations, etc. That, I think, doesn't solve the problem, but shifts the level of granularity.
Great problem, many angles.
For me that's equivalent to the idea that deplatforming isn't problematic, because people can still publish elsewhere or, worst case, still talk to other people.
Key to both ideas is to reject the social significance of operational scale as well as the power dimension of gradual influence.
Practically there is quite a lot of power hidden in the leeway and the bigger a company gets the more problematic their influence is for society as a whole.
The second argument, I think, is that there is nothing problematic about content demonetization because we always can trivially construct a plausible advertising interest against any unfashionable content, hence it's not primarily seen as a chilling effect but something innocent that just, by accident, ends up continuously narrowing the conversation towards the presentable and trivial.
I think this argument isn't great. Just because there's innocent intentions at play it does neither show that there are only innocent intentions at play nor that the overall venture does not, in the end, have bad consequences for society.
If our ad-ecosystem would allow advertisers to nudge a TV station towards what news they show, it would be a bad ecosystem for society, even if it's understandable that someone does not want to show their brand next to real talk.
That's not the case here, however. And I don't think we should be so concerned about a slippery slope that we can't allow any discrimination on the part of advertisers or Google.
By the way, I think I sense an undercurrent of "but that's just stupid" in regards to the objection to the extremely neutral use of dictionary definitions. You haven't made that argument explicitly, but for what it's worth, I'd agree with you on that personally. But that's not my call to make, or yours. (And if I'm imagining that undercurrent, then my apologies.)
More importantly, the issue is that there’s no recourse in these cases. It’s downright stupid that you can report a dictionary for this and get them permanently banned. If the issue is don’t run ads on naught word pages then google should make this list public and stop ruining businesses by practicing “I’ll know it when I see it” style moderation by algorithmic bots without human oversight.
It's a damn definition, not an article or essay on rape. Facts shouldn't be censored because someone feels it might offend their delicate senses.
I mean it's fine to have those pages. Google also ranks them for those keywords.
But I can understand why a company won't allow someone to specifically run ads on those keywords. Like the above comment about ebay talks about.
I'm not sure whether that happened in this case or not, but that was my question
This is the same problem in a different skin: words to not equal intent. When we only judge by words we restrict good faith information and promote bad faith euphimisms that do real harm.
Source: I work in Google. Not in Ads, but this topic has been discussed in TGIF at some point.
No humans, just AI (ML) decisions.
Won't help for a US small business selling on Amazon, I imagine
I wonder what happens if you file e.g. in small claims court against a company like Amazon? They'd probably never get the message, and even if you win due to them not showing up and making their case, good luck enforcing the judgement.
Short of hiring a major law firm whose letterhead might get someone's attention and/or making major waves, I don't see how Joe Sixpack can force a human appeal without major monetary outlay.
I think their strategy of burying their head in the sand and just ghosting you works probably pretty well for the large majority of cases where people simply won't bother (or be able to bother). The cost for the one or two cases that have the energy and commitment to fight is comparably minor and quickly resolved once it hits the front pages somewhere.
Small claims court seems to be a US thing so that doesn't apply in most areas where GDPR is available anyway.
I doubt amazon would just ignore a lawyer's letter. If they do, enjoy your free income I guess
In GDPR cases, you don't go to court, you go to your local data protection agency (pray you don't live in ireland). They will contact Amazon and ask for a statement. Ignoring them isn't something you can do if you value your revenue, because unlike single customers, the agency's task is to go after you with a problem for years, ghosting doesn't work, and they can issue legally binding fines you can't ignore.
Enforcement of either a court order or fines is easy if you got the title on them. If they continue to ignore you can probably get an attachment order, ie, you plus a court official plus some police officers arrive at the nearest amazon HQ and will demand the fine to be paid or they start taking company property to be auctioned off to pay for the fine.
Alternatively you can have a letter delivered by court service, which also is hard to ignore because the court service will require someone at amazon to be read a cover letter, followed by a signature and rather official stamp. After that your letter is considered to be proven in content and delivered. That is something that raises a lot of red flags in legal departments.
https://abcnews.go.com/Business/bank-america-florida-foreclo...
This is ultimately their weakness. Whether it's the binding arbitration exploit that Uber had to deal with or small claims court default judgments these organizations are highly susceptible to coordinated and distributed actions in the real world.
You need to view this as asymmetric warfare where you're using your opponents advantages against them. If they're bigger then you swarm them with small entities. If they can avoid dealing with the public by using AI intermediaries find venues where they simply can't and repeatedly pressure them there.
"Don't struggle only within the ground rules that the people you're struggling against have laid down."
To that last quote, you can often intuit when someone has taken this strategy as the Goliath in the situation will start using terms like "proper channels" and "cowards".
https://web.archive.org/web/20111210150756/http://www.naples...
The company "should" have a DPO, whou you can complain to. If that fails you can complain to the DPA [0]
[0] https://ec.europa.eu/info/law/law-topic/data-protection/refo...
Both enables the aggrieved small biz, and by making it easy, creates a potentially massive problem for the Amazons & Googles who make ignoring the toxic effects of their system a way of doing business
Google accounts have enough worth & history associated with them that they should be able to create some kind of appeal process whereby if you jump through the right hoops proving identity and such, you could eventually reach a human who can intervene?
It feels like they're religious about the idea of having an algorithm decide everything. Works pretty well for some things, but they sure do burn some customers/clients pretty badly along the way for other things.
You’ve hit on the actual problem. This is why so many believe the real fundamental problems the culture is facing with tech companies start with antitrust enforcement.
Skynet's greatest achievement was to make you think it's not real.
Some folks here are in touch with it though. Perhaps you'll have a better luck using HN as the google support channel.
Had my API key revoked because it hadn't been used for 90 days. They provided a link to reclaim it, yielding a LONG form where most questions were only applicable to businesses and service providers.
I answered to the best of my abilities, requesting them to give me access again,also exhausting other available communication options on the matter.
That was about 6 months ago, all I got so far was automated emails telling me it's in review.
Why should Google know that I'm looking at the translation of sexual words in Japanese? Actually, even Google does not want it.
Certainly there are for-pay services as well, but the absolute vast majority of people prefer to pay with their privacy than with their cash, to the point where starting a for-pay service of that type today without a known brand name behind it is not going to pay your hosting bills let alone put food on your table.
I hate ads and am a privacy advocate. However, I'm not going to stretch the facts. Yes it's technically possible to hand back the searched words to ad providers. But as of writing, you have no proof that OP did that. Using adsense doesn't mean you're plugging in that data back into Google. What reason does the app developer have to do that, unless they were paid or something to do so?
"Why would a dictionary need ads" well, because someone made the app. And maintains it. And provides it for free in exchange for a tracking cookie. Since the datasets are most likely free, they not possess the ability to even sell it. Who knows, you certainly don't unless you're the author or did the research.
Privacy and security are not absolute. It's a model. It's a series of tradeoffs. To parrot "but m'privacy!" is only doing harm and isn't solving privacy related problems.
One of the problems with very large corporations like Google is the marginal benefit of each additional customer is negligible. They have so much money that they don't really need extra business and would rather pursue other objectives. Look at all the stuff getting cancelled lately for whatever reason. The whole thing takes place extra-judicially and is not in the monetary interest of these companies. Stakeholder capitalism if you will. I think the old profit driven system was better because at least you knew what the rules were. Now, the way these big companies make decisions is largely opaque and based on secret rules and arbitrary decisions often based on nothing but whim, or worse, a "good enough" algorithm.
As long as there is healthy competition, there is enough incentive to do better, and a customer can switch services without experiencing significant hardship. However, with a lot of the big tech companies, that is no longer the case, and once you run into a problem there is neither free-market nor regulatory mechanisms that can help any individual solve their problem.
Currently, the big tech companies can eschew responsibility by claiming the rights of private business -- having very strong autonomy over how and with whom they do business -- and despite them being de-facto public utilities at this point, the claims are made that regulation is not necessary / not allowed / not productive / harmful. Why this is accepted is puzzling to me, as it is very commonly known macro-economic theory that the usually claimed self-regulating free-market mechanisms no longer work for when large imbalances, eg. monopolies, exist. We have the government explicitly for this reason, to create a counter-balance against forces so large that any individual cannot deal with on their own.
And while it is true that we can expect regulatory intervention to be difficult, especially as the political process cannot possibly keep pace with technology, it is indeed very strongly preferable to giving free reign to private entities with explicitly anti-consumer interests (eg. a corporation has to make their shareholders money, not protect their customers or the environment).
So what's the solution? Maybe regulatory agencies need to be given more teeth, including funding and updated charters. Yesterday's Cartel is today's Big Tech.
An imbalance of power because of employing AI with suboptimal results, but not caring about the negative externalities because, well, they're external.
1984 wasn't conceived from a vantage-point that exists now; the future from here has the potential to be far, far worse. These threads needs to be made into a novel and a movie to infotain people what dystopian concerns are tangible through unrestrained or under-regulated deployment of technologies on the horizon. China is already doing pre-crime and panopticon surveillance.
Maybe the downvoters mean "pro-consumer" in that they're pro manipulating people into consuming their advertisements and incredibly invasive tracking?
Algorithms are not even close to substituting for human customer service, human moderation, and human judgment.
Maybe this means web-scale sites like Amazon's marketplace aren't viable. I'm fine with that implication. Stores like Target and Walmart seem to have the same prices with human-curated inventories, and I don't really miss the millions of extra products Amazon has.
> Maybe this means web-scale sites like Amazon's marketplace aren't viable. I'm fine with that implication. Stores like Target and Walmart seem to have the same prices with human-curated inventories
Not only are they viable, they are hugely more profitable without the human expenses. This inhuman model - yes, I think that is exactly what it is - will displace everything else that simply can't compete in runaway capitalism.
Of course, that's not what we've got...
Given that's possible with human moderation too (cf tax codes), what makes you think AGI will solve the problem?
I don't believe this. The reputation loss alone will cost you a lot of potential money. Stadia is basically dead because the Google brand is completely untrustworthy. B2B customers want to avoid Google as much as possible because it's just a liability that threatens to ruin their business.
I don't expect Amazon to go anywhere, even if people's impression of drops.
The reason it isn’t sufficient to be human-level is because individual humans have all kinds of biases and blindspots which scammers design their scams around.
I’m kinda thinking it may be a necessary prerequisite, as scams could involve anything a human wants, but it isn’t enough.
I agree with you, but what I meant by "viable" was more from a consumer perspective.
> This inhuman model - yes, I think that is exactly what it is - will displace everything else that simply can't compete in runaway capitalism.
While I think you're mostly right, Amazon is eventually going to burn enough buyers that they go back to the more trustworthy brick-and-mortar stores.
Most people I personally know have already done this. They may search for something on Amazon, but they will go and buy it somewhere else.
I'm obviously part of a very specific subset, but it's going to hurt Amazon eventually, especially if the US takes a more consumer-friendly regulatory approach than it has for the past few years (which it will).
I'm sure most people are generally happy to wait 2 days for something to arrive as well. Honestly, it almost feels like Amazon is trying to become an Alibaba and potentially look to take them on with the level of product quality that is now available there.
Half the results on any given product I search point straight back to Alibaba if I image search.
Sometimes I check if Aliexpress has it too, and if I'm not in a hurry I get it there for a huge savings.
They combination is what differentiates them
But the market is driving more and more money/resources to those companies/services.
Let alone Amazon/Youtube/Google are very different businesses, you can say Google has non-existent customer support, but at least for consumers, so far my experience with Amazon is more positive than negative.
I don't dispute that. But a lot of other people have issues with dropshippers, fake products, radioactive products[1], unreliable reviews, etc.
When I stopped using Amazon, I realized I was returning ~90% of all my purchases just because the product was fake, described wrong, low quality, or just not durable enough to last even a month.
Yes, it's great to be able to return things, but it's a lot nicer to feel reasonably confident that the person who sold a product to me actually believes it's a good fit for what I want as a consumer.
1. https://www.japantimes.co.jp/news/2017/09/20/business/nuclea...
Soon, your appeal will be replied to by an AI, too. That way, most people will have the impression that they were reconsidered by a human and found in fault, which will likely make a large percentage of them give up. We're stonewalling real humans by building fake humans :(
And that means GPT could possibly reduce support costs for Amazon.
I don't think this is cynical. It's discussed openly. They want to provide the tools to enable all companies to automate customer service, which is where Facebook's big chatbot push came from.
And that's by design. They don't want to talk to you and help you. You're going to buy there anyway and we all know it.
If your human support team made a mistake, there was an appeal process. If your AI makes a mistake, sorry, bad robot, deal with it.
While people talk up Googles ai (at least they once did) the truth is while they may put out some good research now and again most of the stuff they deploy is crap by design - they can’t afford to spend a dollar per day per customer to run a complex algorithm on your data, they can only spend fractions of a penny.
This severely limits the quality of the ai deployed by large consumer oriented megacorps.
On the other hand you look at some of the ai being deployed in the b2b space or anywhere there is a real exchange of money and you’ll see some pretty expensive and interesting ai research and deployments.
For example, I currently have a company that I technically manage. It has one employee who does maintenance on the infrastructure (AWS Elastic Beanstalk), database (AWS RDS), and code for five different products. The only other people who touch the code are security auditors.
That just wasn't possible 10 years ago. Managing servers alone for those products (~50 EC2 instances, ~10 load balancers, 10 database instances) would've been a full-time job.
It was a non-trivial exercise, and to the best of my knowledge they still have some physical servers, but slowly but surely almost everything moved to Amazon.
Those physical server operations employees slowly left for other jobs.
They company kept growing and employing tech people though. Just tech people doing other things.
Algorithms are human judgement.
In a strict code sense an algorithm is a way of encoding a set of assumptions that you decide on when you're developing the code to run it. Even in unsupervised machine learning an algorithm is the result of human judgement because someone decided to delegate the decision making process to a machine.
Every time a computer makes a judgement call there is a human who is responsible, and should be accountable, for that process existing.
To pull it back a bit, "human judgment" in my usage meant "input processed by human using a brain" and algorithm meant "input processed by machine".
Given those premises, the argument that "algorithms are human judgment" is untrue both by definition and by observation.
For example, when I ask Google/Siri/Alexa a question like, "What is the best way to set up my router?", I will get a wildly different answer than if I ask a GeekSquad employee.
I don’t think the answer is humans or algorithms, the answer is organizations incentivized to solve customer problems. Google doesn’t care about customers, if they did, they would fix this. Customers are locked in because where else will you go.
Just like Comcast doesn’t care because there’s only one cable company, Google doesn’t really have any AdWords or Adsense competition.
I think the solution is more competition so use ddg. Once customers have mobility, Google will work to retain them.
-Someone, somewhere in SV
It's harder to step on a landmine with Google, but Google's landmine is far more devastating.
We've already seen lots of news stories about algorithms shutting down people's Gmail accounts, developer accounts, YouTube channels, etc.
I personally went through 6 months of hell when my brother's Gmail account was locked due to "inauthentic behavior". He had become disabled and unable to log in, and when I logged in from my machine, Google's algorithm flagged it.
I went through an infuriating loop of trying to unlock it and hitting the same algorithm. There was no number to call. I would've paid $10,000 to get back into that account, which would've required 10 min with a customer support rep.
I tweeted at Google and they told me I had to log in to Twitter with his account in order to get help (obviously impossible for me). It was the most angry I have ever been at any company. It locked me out of important medical documents, family messages, social media -- everything.
I finally fixed it by asking a friend who was a marketing exec at Google to help. He said there's an internal form you can use for this situation -- only available to Google employees.
Other people have had similar experiences and had livelihoods destroyed.
Whatever you think about Comcast, they will never ignore you and let your entire digital life get locked away without any kind of recourse.
This is an enormous advantage.
On both of those sites, you can change the filter to show only products that are available in stores.
Those in-store products still have whatever guarantees they did 20 years ago (selected by a human working for Target, probably going to be pulled off shelves if they result in too many returns).
People use them because they are cheap and convenient though. It's the same with algorithms. If you charge margins high enough for good customer service you will get out-competed by those that don't on price.
They may be great for 80% or even 99.9%, but they cannot be 100%.
I guess going to Target is much easier to get the necessary shopping done.
Salomon, Nike, and Garmin all got direct purchases from me recently because of what you describe.
I've abandoned Amazon because of that, until they finally add a "no third party sellers" option. And also a "don't show items we don't actually send to your country" option.
I'm a technically proficient developer who's very careful about where I spend money online. If you're not very internet savvy, it'd have been really easy to miss that. In the end I've ordered from a national chain retailer which are doing click and collect nearby.
- Buy shoes from the Addidas website and you can have your name printed on your shoes for free,
- Buy from Quantas or Ibis instead of a price comparator and you’ll have free cancellations or rebooking for a period.
That you know of. There's really no way to know. Counterfeit products could even be manufactured at the official facility, just with no (or even failed) quality control.
https://www.theatlantic.com/technology/archive/2019/01/count...
Doesn't exactly remove counterfeit products, but goes a decent way towards correcting the fake review problem. Once you've crossed that bridge, you can focus more on the top results without paid bias skewing them.
Searched around online and found it mentioned a few places, but ultimately it looks like they collect sales and order info and resell it... Damn. I'd be happier to pay a small fee instead.
Details on the info they collect and how they use it is in their privacy policy: https://www.fakespot.com/privacy-policy
Then Amazon would be wary of automatic bans.
Of course this must apply to every site like Amazon.
I am a bit down because I’m not aware of any success in fixing this kind of problem through regulation.
I’m trying to think of a regulated industry or company that has service as good as even Walmart and can’t come up with anything.
The hope would be that at least they have some bank level of regulation where they at least have to respond by snail mail and stuff.
Similarly, when you “don’t have a choice” except to work with, say, Google, there is no incentive for them to provide good service. Find a way to introduce competition and everything magically transforms.
The existing anti trust rules are enough to push these monopolies to stop the anti competitive activities. Just like anti trust rules were used against Microsoft many years ago, which opened door to online competition like Amazon, Google, Apple etc.
EU (and UK) citizens have the right that automated decision making that affects individuals must:
* be documented, and individuals must be informed about it.
* include simple ways to request human intervention or challenge decisions.
* be regularly checked to ensure it's working as intended.
This applies to anything with "legal or similarly significant effect on individuals". As examples, the ICO there has "automatic refusal of an online credit application" and "e-recruiting practices without human intervention".
IANAL, but I imagine this would cover the OP's situation if they were EU-based, and most similar complaints (Apple unilaterally pulled my app down incorrectly and won't answer my emails, Google wrongly closed my Gmail account). If a lawsuits & penalties start appearing under that umbrella, it might help change the monopolies positions.
Doesn't help in the US of course, but hopefully they'll follow suit in the coming years, which would apply quite a bit more pressure here.
But other important Amazon marketplaces just won't respond, or don't follow up. I Had the UK tell me if I got a VAT they would enable my account; I did the months of work required to get VAT registered and then responded to UK support that I was ready to go, and have had all of my emails (~1 a month) ignored for over a year now. Why tell me to get VAT and then just pretend I don't exist? They just don't care. Even worse Amazon will INVITE me to join certain marketplaces, which I have to pay to join, to only later then find that I'm randomly blocked in that marketplace as well. Oh yeah, no refunds either.
I've escalated to jeff@amazon and had multiple people there tell me they're in the middle of unblocking the account everywhere, and weeks go by and nothing happens, and then they too just stop responding my emails. At this point jeff@amazon doesn't respond anymore either, even though I made certain to not repeat my request more frequently than once a month. It is like hair-rippingly frustrating and it is the biggest stress point for my business currently. The worst part is this is my main source of income so I'm just relying on their support people who are wholly unaccountable.
I wish I had options as to otherwise but Amazon is such a behemoth now that most of the customers who were purchasing through my own ecommerce store a few years ago now demand to purchase on Amazon or not buy at all. I'm losing a ton of money because of this and have absolutely no way to get help. I cannot even use words to convey how frustrating it has been..
Some people seem to think no web market is better and we should go back to brick and mortar mega corps like Target and Walmart, but good luck getting shelf space there.
p.s. Can’t wait to try the drugs.
It probably *is* 99% accurate - there is a lot of fraudulent activity online.
That 1% false positive rate really hurts though, destroying people's livelihoods.
I think that people developing machine learning models sometimes don't think about the consequences of the false positives enough.
There is a great section in David Spiegelhalter's book The Art of Statistics which got me thinking about this.
That said, in many other applications of AI, a 1% error rate is probably a huge success. AI does not need to be perfect to be considered successful, it only needs to do better than a human. In this example, AI failed miserably at that, but there are plenty of applications where AI errs far less frequently than humans do at the same task.
def is_fraud(x):
return False
You'd get an email, "Hello, there might be a problem, here are the details" and if you made some good faith attempt to reply, be able to talk to an actual person, who could potentially escalate your issue.
I know our salespeople had constant troubles with the U.S. customs because we always had to declare whether our devices were medical equipment or not and different TSA agents had different opinions on how it should have been filled in. They often got upset regardless of how the form was filled in.
Legally speaking it is not a medical device, but some agents disputed that because it was used in the medical field at hospitals. Other agents were upset for wasting their time if you did declare it as medical equipment.
The real TSA agents however, do appear to care. And I'd also argue they're also quite competent. But Snowden was indeed correct that 95% of the people you deal with at the "federal government" are contractors - of varying quality.
What you’re describing is something we run into with the US specifically quite often. It’s kinda crazy, but the US has really become a stagnant market; nationalistic economically, and difficult to do business with unless someone gets to take a big piece of the slice, and very conservative when it comes to innovation.
I would really like to talk more if you’re up for it.
For now...
This forces me to use smaller (often local) alternatives, and honestly I don't miss anything.
So my 2ct: Stop complaining, instead do something. Stop using Amazon, Google, Ebay, Facebook and the likes.
YOU are the one that makes them unavoidable, so just stop doing so.
I ordered a book, they sent the wrong book. They argued over whether it was the right book (they had different isbn), they made me buy an envelope to ship it back and only refunded my money once they received it and cleared it.
So I was out the price of the book for three weeks.
Old Amazon wouldn’t have screwed up, would have mailed out a replacement book immediately, etc.
They seem to be stupid in handling exchanges and cheap now.
This doesn't seem obvious at all. It could be any number of other malicious actors. Off the top of my head: an unsatisfied customer, a troll, a disgruntled former employee, etc...
Put another way: if a single unsatisfied customer is able to take down a product then there must be dozens or hundreds of profit motivated take downs. The article is not naming an exact competitor, but dismissing the accusation as "it could be an entirely different random person" does not change the issue's substance: Amazon is broken.
Not sure what the intent of this observation is - the same is just as true: 'if a single competitor is able to take down a product then there must be dozens or hundreds of profit motivated take downs'.
> The article is not naming an exact competitor, but dismissing the accusation as "it could be an entirely different random person" does not change the issue's substance: Amazon is broken.
Considering there are only so many no-buckle belt vendors, stating that it obviously a competitor casts aspersions on these other vendors without any evidence. But no, it indeed does not change the substantive critique of Amazon.
This seems likelier since, as you say belt manufacturers are few, but blackmailers can go for anything anywhere.
Big tech companies are full of people acting rationally, seeking to maximise metrics that are easy to track and sound good, but have negative externalities that are hard to see or to measure.
It is a bit dystopian, but you can already see some of the adaptations, such as the very existence of SEO and its significant use by businesses.
There are already categories of items I don't ever try to buy on Amazon: farm equipment, fencing supplies, pet food.
And I'm currently looking into email solutions that won't leave me completely screwed when Google arbitrarily decides it's time to delete my gmail account. There's just too much associated with my email address for it to be in the hands of someone who won't answer the phone.
IMO, these companies are winning because they're better than what came before them, but they're far from the perfect solution.
This sounds an awful lot like the experience of the Terraria developer banned from Google earlier this week.
- Seller is banned by an AI agent for arbitrary reasons. There is no reasonable phone recourse
- Seller tries to make a _human_ point in a blog post hoping it goes viral
- The blog post is noticed on different platforms and buzzword of AI wrong doing is amplified to a certain level
- If some arbitrary threshold of _human_ reaction is noticed on the social media, the company in the wrong will probably do something about its AI bot going haywire
Thus, the bot's common sense holes will be covered by collective wisdom of the crowds (which maybe, just maybe, will be integrated in the bot's training loop).
If no human reply is given, it starts to exponentially pile up the fine, until the issue is solved.
They shouldn't be able to allow people to make their livelihoods on their platforms and suddenly cut them off without even giving proper support. They should pay for this - innocent people go through an immense amount of stress and get stripped off their money and the product of their work.
If the answer is: "they're too big to have humans manage this", then you know what the reply is - cut them into pieces, half, fourths, who cares. Slap a user account cap on that shit with a number that a human workforce can handle.
I feel no sympathy for growth problems of this massive companies. I really hope European Union starts to crack on this shit, and hard.
For now I think that voting with your feet is still a recourse, complemented by vocal messages on social media to maximize hit-back.
This has companies make an effort to comply, or other companies refusing to comply and they exit the market not providing products/services to EU citizens, and that's fine!
That's why I believe regulation would do wonders here. Imagine Youtube, or Amazon, have to make up for the loss of revenue of someone they mismanaged and refused customer support? Sprinkle that with fines, and it would start to shift the whole "AI for customer support", which simply doesn't work.
Regarding the cutting into pieces, as long as the user base is cut into manageable size. Amazon has several user bases: AWS users, Amazon Buyers, Amazon Sellers, etc.
In this case, Amazon Sellers are treated like shit, literally. So maybe Amazon Seller platform should be spinned off and have an independent structure dedicated to it. Because currently there's contrafaction, fraud, and a lot of illegal things happening while they give no proper recourse to the damaged parties - unless you're Nike, or Nintendo.
>Remember that lobbying for antitrust splits will be pursued by someone with enough political ambition to offset the pressure in a long run. What would be their benefit?
In this case, for the whole European Union project to start to walk the talk, they have to start to make visible changes. Else the credibility of the union will start to crumble.
The idea of the small claims court is that you don't use a lawyer, so there are no complicated forms to fill in, and if it goes to court you just present your evidence in front of a magistrate in plain English with no legal mumbo jumbo.
Starting a case in the small claims is usually a splendid way to force your way past the army of bots and AI customer service shields, and get your complaint seen by a real person. It usually gets things resolved very quickly. But not always... here's a good story:
Last year I was overcharged £6 by Opodo on the credit card surcharge for a flight booking. I spent two hours going round in circles with their Facebook Messenger automated customer service bot. There were no other contact methods given. So I took them to court, claiming:
Claim: £6.23
My time interacting with chat bot (2hr 22min) @ £100/hr: £236.67
My time preparing claim (1hr) @ £100/hr: £100.00
Court fee: £35
Total: £377.90
... which I think is extremely reasonable of me. They didn't file a defense, so I won the case by default. Even after winning the case, I still had no method of contacting them, and no payment was forthcoming, so I went to the next step which is to file for a warrant.
This cost an extra £95 but it included a bailiff visiting their office, who would be able to take goods to the value of my claim. Bring it on, I say! So I apply for my warrant and wait.
Nothing happens for 6 weeks (this is in the depths of the first lockdown so I wasn't surprised by the radio silence).
Then I get a call from a very helpful chap who tells me he's the bailiff. He says I probably shouldn't hold my breath waiting for this money because the registered office is just a plaque on the fall with no staff or office... and he has FIFTEEN warrants outstanding for Opodo.
The next step would be a warrant against Opodo's bank which would allow me to freeze their account... but there was no easy way to do this and would probably require a lot of paperwork. So I gave up. My belief in the UK criminal justice system shattered.
Then about a month later I got an email from Opodo:
"We are contacting you from Opodo after receiving the Judgement for Claimant. In order to proceed with your refund of £377,90, ..."
... and within a couple of weeks they had paid me in full! I'd highly recommend the small claims court - it takes literally five minutes to file a claim here... check out: https://www.moneyclaim.gov.uk/
The whole "can't get legal fees" is moot in the US since awarding of legal fees to the winner is extremely rare.
And whats always shocking every time I need to contact a large company via some fast response required, its never available.. Phone calls are passed around department to dept that cant deal with significant situations, email support takes far too long, live chat rarely exists.
It's mind blowing how such large companies don't have these services but at the end of the day it's a numbers game for them.
There just are more people that don't complain and their bottom like is still fine.
Really really frustrating as there is no even recommended customer service requirements for companies. I'm sure in the above case there is money lost for this action by the competitor. Will the OP be reimbursed ? Of course not.
Classic Amazon, competitor most likely has financial link to Amazon, even might be their puppet company.
* methoxetamine, a synthetic hallucinogen
* jimson weed, a literal weed from the Americas that causes severe hallucinations and paralysis
* coca leaves, a traditional Andean herbal stimulant that are the natural source of cocaine
None of these items are related or similar at all, and none of them would be used in making belts. Seems like a difficult error to make through mislabeling!
Also, every time I've had an issue with an amazon order that I've needed a human to help with, I've been able to get one on the phone or via chat.
I suppose there are some Amazon sellers who have specific items that (contrary to the public listing) actually contain drugs. In that case, people could order them as a way of buying drugs that would look like some other kind of retail business from the point of view of Amazon, shippers, and tax agencies. The seller could try to make the listed product extremely unappealing or uninteresting to the general public, or have some kind of code or out-of-band contact method that people could use to indicate that they wanted and expected specific orders to contain the drugs instead of the listed item.
Also
Also, it looks like the "people lying in order to harm, steal from, or extort others" issue hasn't been solved for any side of a two-sided market or review system. :-(
Edit: I was first thinking of saying that this is a way that drug prohibition really empowers random people to hurt each other with false accusations. But this is probably true for any kind of high-stakes accusation. As long as there's any kind of contraband or dangerous item that Amazon cares quite a bit about banning from its site, and as long as there's a significant market in that item, people could make spurious accusations that any seller was secretly trading in it.
It seems that it's more important for Amazon to have the lowest prices rather than being a trusted reseller.
(It's so bad that I've stopped buying from Amazon if possible. I'm just too worried that I end up with a counterfeit product.)
It is unfortunate that these trust based systems are being abused.
Most tech companies are run with as few employees as possible, that is why SV FAANG engineers make so much.
We can look to china to understand a bit more of the statistics and personnel required to effectively moderate a system. e.g. number of censors per number of users.
This means that companies like facebook have 2-3 orders of magnitude fewer employees needed to moderate their system. i.e. they are effectively unmoderated. (this is why it was no surprise that facebook was used to plan the capital riots, even though FAANG benefited by deplatforming parlar for political reasons).
We are in an interesting situation where there are probably similar levels of legitimate and illegitimate takedowns but nothing will change because the people with power and money have what they need.