Clare Corthell

Product Manager

✈ San Francisco

✉ mail / @clarecorthell

Last Expedition: Arctic Ice Road Maps
Last Talk: Fixing Cultural Bias in Machine Intelligence

I lead teams building intelligent products with real-world impact.

[ Essays Talks Interviews Δ Open Source Software]

Algorithmic Harm & Engineering Ethics

When we use data to predict, we carry the bias of that data into our predictions. Predictive applications treat people differently, because of the way we encode, collect, normalize, and model data about people. This is particularly problematic in domains like loan assessment, where historical creditworthiness decisions are used to award a given person's credit limit. I work to enable greater responsibility via public and private talks on applying ethics in Product Management and Data Science.

Open Education & Data Science Education

Dedication to open and free education resources, especially in technical fields. I authored The Open Source Data Science Masters.

Design for Hybrid Intelligence & Language Understanding

Though much of humanity’s knowledge is written down, it largely remains locked inside natural language text. Extracting this knowledge from text will allow us to integrate the talents of machines and people to create more adaptable, intelligent systems than ever. Modeling knowledge and language is one of our greatest challenges in AI.

Integration of human and computer work for greater productivity and combined intelligence requires deliberate feedback system design. Optimal integration of these two essential contributors will unlock the highest possible performance in AI systems.

(c) clare corthell, 2018