4 ms·
One product (power bricks for MacBooks) had over 1000 5 stars, then the 1 star reviews came in where people actually had their power adapters catch on fire. Ama
by sundayedition 7y ago
One product (power bricks for MacBooks) had over 1000 5 stars, then the 1 star reviews came in where people actually had their power adapters catch on fire. Amazon, or maybe the seller, shut the review/product down. Now, it's back:
https://www.amazon.com/KUPPET-MacBook-17inch-Compatible-MacBooks-Produced/dp/B07P56DJLV/ref=mp_s_a_1_fkmrnull_1?keywords=Kuppet+macbook&qid=1555375810&s=gateway&sr=8-1-fkmrnull https://www.amazon.com/KUPPET-MacBook-17inch-Compatible-MacB...
Lots of verified purchases. All the reviews are within the same date range. It doesn't take machine learning or advanced AI to catch this; a simple SQL statement should be enough to flag these. But still, they persist.
- harry8 7y agoErgo, Amazon cannot do Stats/ML/AI/call it what you will. The claims that they have an AI advantage don't pass the sniff test.
- sytelus 7y agoIt's not ML/AI problem. The problem is that Amazon wasn't able to punish counterfeiters financially or legally which in turn may be because they don't have a process to ensure seller identity that is legally traceable. This is equivalent to someone coming to BestBuy, sell to them bunch of fake products and get away with it without a scratch every single time. You wouldn't blame that BestBuy doesn't have AI system, would you?
- shrikant 7y agoIt doesn't actually what the nature of the problem domain itself is, in this case -- you can't solve a problem that you don't want to.