Without commenting on the specifics of this case. This seems like a pretty tricky regulation/antitrust issue! Amazon also creates basics products, which directly compete with their sellers. Does Amazon have an unfair advantage, bc they have data from fulfilling sellers' orders?
One could argue grocery stores have a similar "unfair advantage" since, e.g. Walmart observes cookie sales, and can choose which kind of cookie is more profitable to sell. Is this also an unfair advantage?
But there's much more balance-of power between sellers and distributors in the grocery store case. Walmart sells many cookie brands, Kraft sells food at many different grocery stores
Sellers on Amazon have less choices of distributors. Amazon also has the ability to use "ranking shenanigans" to favor store brands. A thought: Walmart could in principle do this too, putting store brands in more valuable, closer-to-eye-level shelf space!
Perhaps a difference is that (my understanding is) Walmart's distribution contracts with brands specify that products must be put in "good shelf space". The contract specifies the distribution tech, there's less "ex post hold-up"
I wonder why Amazon doesn't sell "shelf space" (i.e. priveleged ranking results) to customers in a similar way? Guesses:
- Market's too immature: will happen eventually, but not
- The AI-ranker uses click predictions, which isn't amenable to contracting on
I suppose an issue here is that Amazon can always _claim_ "hey we can't commit to rankings because it's all AI". But I suppose there's currently not much regulation preventing them from "steering" arbitrarily towards their store brand products
The huge size imbalance between sellers and amazon, plus the opacity of the AI-ranking algo relative to shelf space, means it's very hard for anyone to prove allegations that Amazon is/is not "steering", and even if it is, what realistically is a seller to do about it?
It's an interesting set of issues. I wonder if there's any good comparisons in classic antitrust settings. I guess one similar case is big tech's ability to clone startup apps with cool features (fleets, instagram stories, etc.)
But this case is (to exaggerate somewhat) as if FB could see all of snapchat's user data, clone snapchat, then make it very difficult for a user to ever install snapchat, literally aggressively pointing any attempted installs to FB's clone
Signal-boosting @TradeDiversion's link here: tl;dr is that Amazon private labels are a fairly small share of their sales, relative to chain store private labels
Couple people pointed out that Amazon does sell "good shelf space" - ads! But I guess the difference is, if you have an "organic" search feed and an an ads feed, Amazon could in principle always bump their private label stuff up in the organic feed
Though I guess the situation is somewhat analogous for grocery stores: just bc you pay for good shelf space, doesn't mean the store can't put their store brand product right next to your product...
• • •
Missing some Tweet in this thread? You can try to
force a refresh
I participate a bit in the econ RA market on the hiring side, and I also hang out occasionally on the @Academic_Econ discord and talk with ppl looking for positions. Strikes me that labor markets (probably all markets!) are shrouded in a perpetual N-way "fog of war"
Each participant in the market sees a small bubble of info, that's available to and relevant to them. No single person in the market actually has the big picture at any particular point in time!
Applicants know the set of positions hiring, but not the screening criteria/hiring bar/etc. I know my own criteria and some colleagues', but other schools could be very different. I see roughly who we win against and who we lose to, but only in a small ball of programs around us
Conjecture: there is a decent (though not perfect) correlation between how well an industry/area/etc is doing, and how nice the median person in the area is
My guess at the reason for this is that people tend to be nicer when they are in environments of relative abundance, that feel positive-sum: on average, helping a stranger will be good for myself at some point
When people sense that growth is slowing down, there is not enough for everyone, perhaps the aggregate mentality shifts to be more zero-sum: my neighbor is generically my competitor, and helping them will tend to hurt myself slightly on average
Prediction: in the near future, every luxury good (handbags/clothes, wine, cars, etc) will have an NFT associated, and Instagram will integrate and allow linking NFT's to photos. I post a photo of me with a fancy bottle of scotch, and the NFT to prove I actually bought it
70% of the point of buying luxury stuff is to instagram about it. Now, at the moment, a nontrivial % of ferrari instagrams are "hey look at this ferrari on the street let's pretend we own it". NFT's would solve this
Now, why couldn't I just borrow a friend's ferrari NFT and instagram using it? Well, if I use the code, my friend can't also use it, and again, 70% of the point of the thing is to post it on instagram...
Interesting alternative property rights mechanism: when you buy a bottle of Laphroaig (who seem not to be on Twitter), you get this cool little certificate
When you enter it on their website (which only works on IE, chrome breaks it) you get this big certificate, which entitles you to a square foot of land at their distillery. They now owe me rent of a dram of Laphroaig per year. Yay! I am now a landowner!
Ownership rights are conveniently bundled with survey apparatus for verifying the dimensions of the plot of land, as well as weather-appropriate equipment for fending off attacks from wild geese, stoats, and otters while attempting to access the land
Here's a set of questions. It seems a large % of people believe Tether is insolvent, which I define as: Tether does not actually have enough USD to pay $1 for each USDT, if everyone redeemed at once
Questions:
1. Why is a USDT still worth $1, if Tether is insolvent? Why isn't it worth like $0.9, or the market's best guess at how much USD assets Tether actually holds? 2. Couldn't a hedge fund or someone "attack" Tether to exploit the fact that Tether is insolvent?
Here is what I think about 1. The key is that USDT cannot trade below $1, _as long as Tether allows you to redeem 1 USDT for $1_. The argument is super simple! If ever someone was willing to sell a USDT for like, $0.99, you could just buy it from them and redeem it for $1!
In tech firms, from what I've heard, it's rare for very early stage startups to have many data scientists. You need some product guys, hackers, and sales/marketing people. Make something, try to sell it, pivot if it fails, repeat until you succeed or run out of money
At this stage, I guess data science isn't needed because success or failure is obvious. Or sufficiently obvious not to need p-values. You have users, or you don't. You also realistically don't have big enough N's to actually run experiments even if you wanted to
Data scientists seem to enter later when firms are more mature. Changes are more incremental, and they're run through A/B tests, datasci folks pore over metrics, before deciding whether or not to shift. I think this is an intrinsically slower process