"Why Do Businesses Fail At Machine Learning?" by Cassie Kozyrkov from Google
Vložit
- čas přidán 2. 08. 2024
- Robotex International Conference: Generation R attracted 924 attendees from 54 countries to Tallinn through November 29 - December 1, 2018.
Altogether, we brought together 80 speakers from 23 countries who appeared over 5 stages. This video of Cassie Kozyrkov from Google was recorded on our Big Ideas Stage on 30.11.
Robotex International Conference 2019 will happen through November 27-29, 2019. Tickets are on sale, now!
#Robotex18 #50000Robots - Věda a technologie
Would have been great to see the presentation screen. Nevertheless, amazing speaker.
What so many people forget however, if you want to make Italian food but don't have access to quality ingredients, forgettaboutit! That will be you first problem to solve: The quality of the input data. It if sucks, the output will be a lackluster tourist trap in Oslo.
Cassie Kozyrkov=gentleness, grace and great intelligence....all wrapped together
Her movement is soooo graceful, even the cameraman just got carried away...you can tell!!! Before you call me sexiest, the presentation's GREAT TOO!
There are some fantastic analogies in this talk.
The way she is explaining is awesome. Her soft skills rare also very inspiring..
I love the talk, the explanations, the examples, analogy and the concept behind the presentation, the subject.
Just read your article in TDS and came here. Really great.
A lot of ML professionals are quite arrogant and condescending. The bit of advice about the attitude is one of the most important in business environments
Brilliant!
Susul Tetrabuana Soeryo,Attend for whatching.
She's is becoming my teacher.
really innovative speaker
I like the line 'is your business problem relevant? '... Do you have similar talks for students & professionals aspiring to shift to ML/AL...i see students too getting off track in their ML learning journey...
OMG!! fall in love 😍
When hiring is done by skill less MBAs this type of thing happens.
The nice thing about a microwave is that results are reproducible.
the creepy thing is I don't know if the speaker is a robot or not.
She is a human and is one of the smartest you'll ever meet
Her movements looking very intentional and metered, it's like watching a very slow version of the robot dance.
Keeps using the same bad Microwave analogy in same talks.
She makes a major error in logic in the first 2 minutes. Equating machine learning with a microwave is a false dichotomy as a microwave produces a known output. Machine learning is function approximation. The output depends upon whether a function exists between the inputs and the outputs. If the relationship between inputs and outputs fails to persist, the model becomes useless. If the microwave had to be replaced every few days, people would stop buying that microwave and that company would go out of business. ML on the other hand accepts that models break all the time to the extent that breaking is to be expected.
@@rickjohnson247 I'm pretty sure no one actually think Machine Learning is a microwave. It is called in layman's terms.
@@rickjohnson247 you are right but she is a great teacher nonetheless.
@@rickjohnson247 She was not explaining how machine learning works. She was explaining how to allocate resources in managing the application of machine learning for business success. The fact that you, someone who clearly understands machine learning well, cannot even understand the intent of her analogy, is why she needs to open with that analogy to a general audience in the first place.
She makes a major error in logic in the first 2 minutes. Equating machine learning with a microwave is a false dichotomy as a microwave produces a known output. Machine learning is function approximation. The output depends upon whether a function exists between the inputs and the outputs. If the relationship between inputs and outputs fails to persist, the model becomes useless. If the microwave had to be replaced every few days, people would stop buying that microwave and that company would go out of business. ML on the other hand accepts that models break all the time to the extent that breaking is to be expected.
The point went over your head. It wasn't an apples-to-apples analogy, but just a place holder for people to understand the analogy. Secondly, you don't know what a false dichotomy means, because making a bad comparison or bad analogy is NOT a false dichotomy, which is an either/or fallacy. So, you're wrong, and you're wrong.
@@brotendo thug life XD
@@brotendo No matter how much you agree with her, you're not sleeping with her lmao
@@rickjohnson247 Eww dude. You think she's attractive? Lmfao.
@@rickjohnson247 Wow nice reply. What are you? A teenager who just finished his machine learning course on Coursera? Lol