How bad data keeps us from good AI
1,455,036 views |
Mainak Mazumdar |
TED Salon: Dell Technologies
• October 2020
The future economy won't be built by people and factories, but by algorithms and artificial intelligence, says data scientist Mainak Mazumdar. But what happens when these algorithms get trained on biased data? Drawing on examples from Shanghai to New York City, Mazumdar shows how less-than-quality data leads to AI that makes wrong decisions and predictions -- and reveals three infrastructural resets needed to make ethical AI possible.