Security, Transparency, and Ethics in AI
04:44 — What it means to grow up in a country with an established culture of trust.
09:54 — Transparency doesn’t mean that everything has to be open. It is about creating an environment where stakeholders have the correct information.
17:51 — We must move from establishing the principles of ethics in AI to application.
23:32 — Gender diversity in data science does need addressing, but the area of ethics has already attracted a lot of women.
Meeri Haataja talks about what it means to come from a country with an established culture of trust and how that affects the way Finns are training and implementing AI. She believes transparency in AI is important but that visibility should be determined by need — consumers, for example, will not require the same information as auditors.
There has been much good work on the principles of ethics in AI, but now, Meeri argues, we need to look at applying those principles. She calls for a pragmatic approach to be headed by industry leaders. Importantly, Meeri would like to see this practical application happen collaboratively as this would help create a unified and citizen-centric approach.
On the subject of gender diversity, Meeri notes that while data science is still male dominated, the field of ethics has attracted a large number of women and wonders if we have realized that tech embraces more capabilities than just engineering. She discusses the challenges of creating a diverse team stating that means more than just gender or race it must include culture, career backgrounds, and experience.
Listen to more about AI, ethics, and privacy in our podcasts with Andrea Bonime-Blanc, Joe Stuntz of Virtru, and Laura Noren of Obsidian.