Netflix is not a tech company (2019)

Re:Netflix is not a tech company (2019) by spoonjim on Tuesday September 07, 2021 @02:22PM (#28448070)

There is probably still a lot more AI/ML work to do at Netflix. For example if you are browsing through standup comedy specials and you see the trailer joke, and you skip to the next one, they know exactly where in the joke that comedian bombed for you. So for other people like you, they can show a different joke from their special, and for you they can build a profile of the kind of humor you like. Then if you start a video and stop it, they can know that the joke they showed as a trailer was not representative of the first impression the comedian gives in the special. etc.

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by mulmen on Tuesday September 07, 2021 @02:33PM (#28448213)

It’s an open question if this kind of capability actually has a positive impact on the bottom line.

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by cletus on Tuesday September 07, 2021 @02:47PM (#28448395)

Honestly that's my experience with ML in general. And I know any organization that does a lot of ML has things like backtests and holdouts to supposedly measure impact on key metrics but I've seen too many cases where adding X has a +10% impact in some top line metric and then removing it later (and returning to the initial state) has another 5% increase in that same top line metric.

Maybe we need to revise the famous phrase to: Lies, damned lies, statistics and machine learning.

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by spoonjim on Tuesday September 07, 2021 @03:13PM (#28448654)

They must know whether people encouraged to watch more Netflix actually retain Netflix at higher rates. As their prices have gone up I assume this is becoming a bigger and bigger issue.

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by mulmen on Wednesday September 08, 2021 @02:54AM (#28453930)

How does your ML model control for opportunity cost? How do you prove a causation between watch rate and retention?

Do algorithms that encourage viewership actually increase revenue, or retention? What happens when the algorithm is so good I under-sleep, miss work and get fired, eventually losing my ability to pay for Netflix? What if I see this in myself and cancel my subscription, instead of using Netflix sustainably? What happens if I genuinely watch all programming I find interesting and move on to reading books on the same topics?

In other words, if the Netflix catalog is perfectly sorted by relevance to my personal interests, how much of it will I engage with? How far can Netflix content creators and licensors follow me down the rabbit hole?

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by jakeinspace on Tuesday September 07, 2021 @02:44PM (#28448355)

I don't see how this example would necessarily benefit me as the content consumer/customer nor Netflix. What if the clips that Netflix learns to show me aren't representative of the entire special, and I'm disappointed? And while Netflix to some extent wants to have me spend time using their service (to ensure I feel like I'm getting value and don't cancel my subscription), watching more things I don't enjoy may not help.

[ Parent | on HN ]

Re:Netflix is not a tech company (2019) by jamesfinlayson on Tuesday September 07, 2021 @08:30PM (#28451521)

Agreed - the tile for Brooklyn Nine Nine when I'm signed in is Jake Peralta but for Mum it's Rosa Diaz.

[ Parent | on HN ]