I've started my class this semester on Algorithms, Data and Society with a week of grounding everyone in coding and machine learning. My (always wounderful!) students will read Unreasonable Effectiveness of RNNs and social theory the same class! 
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Datasets already contain our collective biased behaviors
-
Absolutely! But that's a different path to what's happening then what people's mental model of the process is--some sort of implicit bias type bias creeping into the programs.
- Show replies
New conversation -
-
-
Was only mentioning Microsoft’s Tay earlier today.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
An unbiased algorithm trained on biased data -- any measure of human activity -- will produce biased results.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Bias in the training set will bias the learning machine.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
There's also a difference between technical competency, competency in humam behavior, and competency in the intersection of those spheres. Those who write code for big human socialization mediums may not know themselves the bias that is being embedded into their code.
-
*competency in measuring human behavior
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.