Guys, its a joke, a meme. Some guy makes these fake notices like the woman who sweats on your couch, maggots appear and you are supposed to eat them to heal yourself.
amazon spent a lot of money on trying to do this and then found out the technology doesn’t exist and outsourced it to india
But that’s not a bowl. It’s more like a box. No, it’s ok. I’ll get on the call at 10pm.
Some companies do outsource their “AI” to India, but automated checkout tech is actually good enough to be used in production now. A plain white background with separated fruits like this is exactly the environment where it works best.
automated checkout tech is actually good enough to be used in production now
not really.
amazon’s just walk out is the leader in this area, and it came out recently that the bulk of transactions, 7 in 10, are offloaded for manual review in india
amazon of course denied the claim, but so in vague corporate speak, and failed to provide figures to counter the 7-in-10. they also did confirm that they’re scaling back just walk out. i don’t think those things would be the case if this technology worked as they were hoping.
Just because Amazon, king of scams, is doing an AI scam, that doesn’t mean that the underlying technology is impossible to use with minimal errors (it’s AI, it’s made of statistics, there will always be some errors).
Anyways, “just walk out” works in a different way than the fruit recognition in the OP or the checkout machines I was talking about. Image recognition of a discrete item over a white background (or a checkered background) is like, the literal ideal case for image recognition accuracy. This is as opposed to blurry store cameras looking at an entire aisle from 20 feet away and trying to guess what item the customer is taking off the shelf. It’s an entirely different problem space in every way that matters.
Anyways, even ignoring theoretical arguments, I know it’s production-ready because it’s currently beong used in production. There are dozens of stores in Calofornia right now that use checkout machines with a camera that points down towards a plain background “pad”. You place the item on the pad and it selects the most likely item in the store based on what it sees. I’ve seen a live demo of these machines where you take ~10-15 pictures of an item from different angles/rotations/positions and add it to the list of recognizable items, and the machine was able to diatinguish between that item and others accurately. This was in a very candid and scam-unlikely environment (OpenSauce) and by my evaluation this is easily consistent with other known-good image recognition applications.
The Amazon shop is a lot more complicated than a few berries on a white shelf.
not in the ways that matter, and small, organic items like individual berries are far harder to account for than standardized product packaging
Could be or could be the berries are put in the same arrangement each day and it’s just tracking which black blob disappears.
pretty sure items on a shop shelf are in the same arrangement each day
That’s not necessarily true-- in fact, two similarly packaged items that are otherwise different might actually be harder to tell apart when packaged.
which is why just walk out also had rfid tokens on all their products
you can’t do that with a strawberry unless you like your fruit crunchy
[Sorry, double posted, my mobile connection is pretty bad rn]
Just because Amazon, king of scams, is doing an AI scam, that doesn’t mean that the underlying technology is impossible to use with minimal errors (it’s AI, it’s made of statistics, there will always be some errors).
Anyways, “just walk out” works in a different way than the fruit recognition in the OP or the checkout machines I was talking about. Image recognition of a discrete item over a white background (or a checkered background) is like, the literal ideal case for image recognition accuracy. This is as opposed to blurry store cameras looking at an entire aisle from 20 feet away and trying to guess what item the customer is taking off the shelf. It’s an entirely different problem space in every way that matters.
Anyways, even ignoring theoretical arguments, I know it’s production-ready because it’s currently beong used in production. There are dozens of stores in Calofornia right now that use checkout machines with a camera that points down towards a plain background “pad”. You place the item on the pad and it selects the most likely item in the store based on what it sees. I’ve seen a live demo of these machines where you take ~10-15 pictures of an item from different angles/rotations/positions and add it to the list of recognizable items, and the machine was able to diatinguish between that item and others accurately. This was in a very candid and scam-unlikely environment (OpenSauce) and by my evaluation this is easily consistent with other known-good image recognition applications.
it’s AI, it’s made of statistics, there will always be some errors
7 in 10 required manual review
This is as opposed to blurry store cameras looking at an entire aisle from 20 feet away and trying to guess what item the customer is taking off the shelf. It’s an entirely different problem space in every way that matters.
which is why that wasn’t the setup of just walk out
every location was quite literally purpose built with the express goal of making the just walk out technology as accurate as it possibly could be
You place the item on the pad and it selects the most likely item in the store based on what it sees
this is a completely different problem
nobody’s placing the berry or berries they decide to eat or not eat in a separate area before placing them in their mouth
this is a completely different problem
Yes, that’s what I’ve been trying to explain. And no, JWO was not built to be accurate, it was built to be convenient. That’s a very different incentive that will lead to skipping alternatives that are less convenient but more accurate-- like the checkout kiosks I’ve been talking about. I’m not defending JWO and it’s obviously both a harder problem and one that’s not managed well, focusing on optics over accuracy.
nobody’s placing the berry or berries they decide to eat or not eat in a separate area before placing them in their mouth
That’s not necessary, they’re already placed in a nearly ideal environment by the person setting up the berry bowl. Notice how the “bowl” is a white square with each fruit placed in a way where they’re separated by the whitespace. You wouldn’t even need to train a model on the whole bowl, you could just do an image region detection --> object recognition pipeline. The hardest part about the berry bowl would by far be determining the person taking the fruit! (In fact, I wouldn’t be surprised if that was manually reviewed, with that few instances to look at.)
Yes, that’s what I’ve been trying to explain
jwo is a different problem than the separate checkout kiosk you’re describing
jwo is the same problem as is in the image
JWO was not built to be accurate, it was built to be convenient
it was built to be accurate within the boundary of “no checkout step”
at this point it feels like you’re deliberately misinterpreting me
Notice how the “bowl” is a white square with each fruit placed in a way where they’re separated by the whitespace
unless somebody moves or jostles them while taking some fruit
you’re essentially making the exact same naive assumptions about the operating environment that led to jwo’s failures
if “just track which one disappeared” was a valid solution to the problem, jwo wouldn’t have failed
The hardest part about the berry bowl would by far be determining the person taking the fruit
facial recognition is a thoroughly solved problem, at least in terms of the accuracy that we’re aiming for here
That was for automated checkout. Video people counters have been around for years. I’ve worked for companies that used them to count customers by department.
this isn’t counting people. this is working out which item or items people pick up from a shelf and decide to keep, if any. that isn’t just similar to the automated checkout problem: it’s the same exact problem. if anything, this iteration of it is more challenging because a blueberry is a fair amount smaller than a tin of beans.
Good point.
No, facial recognition works (unfortunately), it’s just not good enough to look at an entire shopping cart and know what’s in it lol
If lemmites didn’t have the social skills of a mosquito at a funeral they might be able to pick up on the INCREDIBLY subtle joke here.
I’m not reading that comment chain. What’s the joke?
The absurd premise of the sign?
I thought I missed some joke in the comments.
Lemmites being so divorced from reality they think this is real
Man, everybody is so upset at this, it makes it even funnier.
Welcome to Lemmy!
I was about to say this had obvious plant vibes lol
Here’s the creator: https://www.instagram.com/truewagner/?hl=en
I worry about the state of the world seeing how many people just take shit like this at face value without taking even a second to think about it.
Thats a meme and its funny. Its literally satirical criticism of governments putting “ai” cameras everywhere.
Wow, I absolutely detest people like this. How sad and pathetic does your life have to be to be doing something like this. Why can’t people just be normal???! Imagine if everyone went around spreading misinformation like that guy. Plums are not berries, people! Don’t let big berry claim this juicy deliciousness as their own!!!
Strawberries aren’t berries either.
At what point do we say that science is wrong about berries
You can’t look at me with a straight face and say a banana is a berry.
Look at the whole fruit bundle before they break it apart for shipping. You can’t tell me it’s NOT a berry.
Just gotta accept that we have Berry[culinary] and Berry[botanical] and these variants are unstated and must be inferred from context.
Technically, a banana is the whole big-ass thing that grows on the tree. What you know as a banana is what is called a finger of a banana.
A banana is a berry 😑
Speaking of juicy imagine the shits you would have after eating 27 fucking plums
Omar would know that
World’s most regular Omar
27 plums a month will not have a noticeable effect on your bowel movements.
Imagine thinking plumbs are berries…
This is basically the algorithms of the big tech companies but with extra steps. I guess it greatly illustrates how absurd they actually are and how weird it is to just shrug them off.
The best part? Nobody can agree if advertising ala commercials actually works.
“Nobody can agree”?
Is this kind of like how nobody can agree on whether we landed on the moon or whether climate change is real?
My thoughts too. It definitely works. That’s why I despise advertising and block all attempts at it. Don’t tell me what to think, assholes.
Yeah it’s really bizarre how we just accept that there’s this massive industry dedicated solely to emotionally manipulating people into buying shit. If it wasn’t reality and we just saw it portrayed in a sci-fi movie, we’d think “this movie is dumb, why would people ever put up with that kind of bullshit existing?”
Right? Now we have people claiming you have a moral obligation to watch ads because they support X or Y. Bizarre is a good word.
Anecdotal, but I’ve boycotted products or stores because of annoying ads in the past. And even clever or entertaining ads eventually get annoying if they are shown too much.
I block most ads these days, so longer contribute to the negative ad returns like I used to, but I am curious about how many others like me are out there.
My spouse is like that; they’re so anti-advertising that they’ll close their mobile games once the ad. appears and re-open the game, just to not watch it.
And they’re definitely the type to stop buying from a company because the ad.s were annoying.
Anyone from Canada in the late 90s/early 00s might remember that Canadian Tire guy. His character was kinda a personified commercial, he was just so enthusiastic about the Canadian Tire product that could help this common problem that it was off putting. Even though those windshield wipers that were one curved piece that would conform to the shape of your windshield looked like exactly what I wanted, I didn’t step foot in a Canadian Tire until years after they got rid of him.
That’s the one kind of advertisement I think is acceptable. “Here’s a product you probably didn’t know existed.” That’s actually informative. Be straight up with me, you have products I might need, tell me what they are and we’re done.
The ones that are sneakily trying to make it so I trust a brand with music and high production values are the ones that are disturbing to me. I feel like most companies make both shit products and some good products too. These corporations are massive so it’s not like it’s just one extremely qualified team of people making all of their products. Different divisions different quality of products, so the brand is meaningless. But somehow branding is the big focus of marketing now and it’s all a meaningless waste of time.
Yeah, those ads that feel meaningful but really aren’t.
Though the worst ones are the ones that use emotional manipulation, like making parents think a new minivan or a can of ground coffee will bring their family back together. I think the “play cool music with extreme visuals to make teenagers think we’re cool” also qualifies for emotional manipulation, though it feels a bit less sinister. But the more I think about it, the less I feel like it is more sinister, since they are all preying on complex desires that they imply they will help with but can’t really deliver on.
That sounds like misinterpretation of studies if I had to guess. If anything maybe it would be how well they work at certain investment levels with breakpoints where it is no longer worth putting in additional money, or the efficacy of CERTAIN types of advertising. But IF they work at all? I am highly suspect of that even being a question. Brands that are well advertised are well known, and people who aren’t bothered to do research into every single product they buy are more likely to buy a product they have heard of before.
And if you’re an independent creative, like author, artist, musician, developer, and you DON’T advertise, you will sell zero copies of your work.
Same landlord: follow me on my socials to know more about how your average poop size and weight compares to your neighbours.
I could DESTROY that competition.
Dude you don’t know who you’re up against!!
Poor landlord must go through so many toilet cameras because of you two … the suffering might even make him appear human!
Who uses a poop knife? You’ll be surprised.
Too bad he ain’t on Lemmy. But thanks for the source.
Indeed, just edited in a warning :)
Someone who has an IG account, consider commenting to encourage fediversing up!
Fuck yeah, come join us! Meta sucks!
“nah, we don’t need privacy laws”
“Definitely safe everything’s okay rectangular shelf with latex paint bowl.”
Random “surprise paint”, perfect for placing acidic fruits. I don’t see why one would have a problem with it.
It really being latex paint with toxic solvents is the best case scenario here.
deleted by creator
Meh. It’s a bit less creepy than any website that tracks you, but in this case you get to see the results. Plus: free berries are nice.
Fake - if this was real there would be a category ‘most times pooped in the box’.
I bet Omar isn’t even trying. He could easily eat another 27 plums no problem.
Yeah I’ve been disappointed with Omar’s performance this year, he’s gotta pump up those plum numbers
Dave, if you aren’t in the top ten next month I will raise your rent.
Fuckin freak behavior, what a fuckin creep
It’s okay the law says I’m allowed to do this.
-person who all laws cater to
