|
|
|
|
|
|
by visarga
3533 days ago
|
|
|
Systematic analysis of AI biases is certainly needed. We train models on data, but how is the data collected, and how biased is it? At least in AI we can compensate for biases, but in human society they are much harder to counter. There's hope for a better future if we can make fair AI. |
|
|
"As another example, a 2015 study9 showed that a machine-learning technique used to predict which hospital patients would develop pneumonia complications worked well in most situations. But it made one serious error: it instructed doctors to send patients with asthma home even though such people are in a high-risk category. Because the hospital automatically sent patients with asthma to intensive care, these people were rarely on the ‘required further care’ records on which the system was trained."