March 31, 2021
Ethical AI Part 8: Equality
We provide the same high level of services for all languages and dialects in the world.
Equality. Transparency. Humans first. These are just a few of the principles commonly listed in fashionable ethical frameworks for artificial intelligence (AI) as released by different organizations.
There’s nothing wrong with these AI ethics frameworks: they work well for marketing campaigns and for guiding the technical personnel at R&D agencies. The problem lies in specificity. How can these nice principles be implemented in practice? And how can we make sure they don’t give organizations a false sense that they’ve made their AI entirely free from potential problems?
Back in 2019, when Utopia finally launched its Ethical AI Manifesto, the company had already been implementing all the principles from the launch date of Utopia AI Moderator back in May 2016. The Ethical AI framework was not about creating something new. It was about writing down the Utopia AI way, finding the words for how Utopia’s experience of putting ethics into action.
Equality is a noun. One of its definitions is “the condition or state of being the same in number, amount, degree, rank, or quality.” For Utopia this includes providing the same high level of services for all languages and dialects in the world. It’s an easy promise to keep, since Utopia AI is language-agnostic.
Unequal access to necessary AI services – for instance, to machine-learning-based text analytics – will generate huge disparities between companies. The innovation achievement gap in the field of text analytics and other language technologies can also have a corrosive effect on smaller languages, shrinking their influence while the big languages keep growing.
Naturally, equality should also be considered as equal treatment in analysis of chat messages, news comments, reviews, claims and other sorts of user-generated content. Utopia implements that form of equality by simply not agreeing to build any AI models with prejudice. Utopia AI analyses all items solely according to content and context, no matter who wrote it.
Ethical AI part 1: Time to talk about responsibility
Ethical AI part 2: Powerful tools need a solid base
Ethical AI part 3: Human rights breach as grounds for termination
Ethical AI part 4: AI moderation and freedom of expression
Ethical AI part 5: Values in order
Ethical AI part 6: No prejudice
Ethical AI part 7: Honesty