AI Is Not Biased. Our Data Is.
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
Claire Roberts has spent twenty five years in technology strategy and transformation, most of it as one of the only women in the room. She got in without a degree, on the strength of being more useful than anyone else in the building.
In this conversation with Jenny Garrett OBE and Leah Garrett, she talks about what she saw when a major AI engineering company quietly let its diversity commitments slide once the external pressure changed direction, why researcher Sarah Weier's experiment on how models answer questions about women should worry us more than it does, and why the uncomfortable finding is not about the technology at all. The bias is not programmed in. It is in the record we left behind, and the model is reading it back to us.
She also makes the case for AI as a thought partner rather than a replacement, in healthcare, in engineering, and in the ordinary work most of us do every day. Her instruction is simple and harder than it sounds: don't outsource your thinking, use it to improve your thinking.
In this episode:
Getting into tech without a degree, and what usefulness actually buys you
Being the only woman in the room for twenty five years
What happened when diversity commitments met commercial pressure
Sarah Weier's experiment and what the training data revealed
Using AI to sharpen judgement instead of replacing it
The AI for Equity podcast, hosted by Jenny Garrett OBE and Leah Garrett.