Did Zillow fail miserably due to forcing application of Machine Learning ?
They have some ML models to guess prices. I feel it's another use case where ML is forced just to justify their ML headcount.
No. ML models actually predicted reasonable prices. Issue was due to folks overriding these estimates to meet their quotas.
Thats crazyyyyyyyyyyyyyyyyyyyyyyyyyyyy… You’re likely a ML engineer over there 😊
Not ML, just a regular SWE. ML for pricing home is pretty trivial. Most ML books use this as an example. Tensorflow used to have a tutorial on this. If people don't follow the price recommendations, then what's the freaking point of it
ML for housing price not possible without good uniform individual input per house Bigger problem was the real estate buying and selling fees
Zillow’s ML is only if statements
then what are the ML engineers doing there ? or are they all clowns like in most ML teams ?
It's much more than that, I promise :)
House of cards. Finding out ZO ML director had 0.5 YOE after biology BS. Mfw.
I don't think it's fair to single out a single person. Many of the director+ were held by pip'd Amazon folks. What did you expect?
Yeah list goes on and on. VP with 3 YOE after MBA out of no name school no name company. Another VP with 5 YOE all at Zillow. Amazon? Overqualified.
lol do people really think ML can predict house prices in the future?
How about Facial recognition and fraud detection?
Yea ML is completely useless. Agree. Fucboi is smart though