4 ms·
As creepy as this is whenever it happens, I remain convinced it’s observation bias. Because of how often it doesn’t happen. Quoth Richard Feynman: > “You know
by Swizec 4y ago
As creepy as this is whenever it happens, I remain convinced it’s observation bias. Because of how often it doesn’t happen.
Quoth Richard Feynman:
> “You know, the most amazing thing happened to me tonight... I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!”
You were gonna see ads, or news stories, or whatever. You happened to see this one. And it happened to connect to a conversation you were having. Amazing! What magic!?
Well how many other ads or other times didnt it connect to a conversation and you just don’t remember because it didn’t feel special? Probably many more times than it did work.
Lately I’ve been trying to find alternate explanations. Hm we were talking about going to a restaurant tomorrow. Why is my girlfriend getting all these restaurant ads on instagram all of a sudden? Oh right, because it’s Thursday, we live in a city, I searched Google for “Good Friday restaurants for a date” during our conversation, and we live on the same IP.
- pclmulqdq 4y agoGoogle probably gets higher CTR by showing you many different ads that are oddly specific rather than ads that are more general. Once in a while, they will get one right, and it will feel creepy.
- snowmizuh-04 4y agoI love the Feynman reference. Thank you. And this may be right. However, I also am reminded of Nassim Taleb's Fat Tony, who when asked 'what are the odds of flipping a coin heads 10 times in a row?' responds with 'fuggehdabowdit. it's a hustle.' There's a scientific response, which can be very naive in some ways, and there's the Fat Tony street analysis. As I've gotten older, I tend to value the latter.
- Swizec 4y agoOh I’m not saying they’re not trying. I’m more saying that they aren’t as good as it sometimes looks. Humans are very good at finding patterns in happy little accidents. Much better than the ad networks are in making those accidents happen. If the tech was as good as they say, we’d never ever see an ad that wasn’t a hyper relevant instant click. And it seems unlikely that Apple is selling my iMessage conversations to Facebook. The “we searched for it, created an explicit signal, and forgot because it’s such a habit” explanation seems more likely. Plus if a simpler method like “hey your location is at a restaurant on most Fridays, why don’t we start pushing restaurant ads on Wednesday” works well enough, why wouldn’t they use it?
- orangecat 4y agoAs I've gotten older, I tend to value the latter. Definitely a good perspective to keep in mind. But at Facebook scale, if a billion people flip an actually fair coin 10 times, around a million of them will get 10 heads. It would be understandable for those people to conclude that the coin is biased, but they'd be wrong.
- FabHK 4y agoRight. And 0.1% of them, 1000, will post here on HN how this can't possibly be a coincidence.
- tom-thistime 4y agoI'm sorry, but the Feynman reference may be a little misleading. With respect, it is difficult to imagine Richard Feynman putting "I sent the information to the recipient via the internet, but I hope they didn't read it" in the same category as "no way they could know; must be telepathy."
- therealdrag0 4y agoIt’s not “I hope they don’t read it”. It’s: — they say they don’t. — developers who work on it say what they’ve worked on doesn’t. —dispute this conversation being a decade old, no one has reverse engineered the client or network traffic to prove it does afaik.
- unity1001 4y ago> Because of how often it doesn’t happen. So sweet boxes from Kazakhstan suddenly appearing in your ads is due to observation bias...