Welcome back to Voices of Trusted AI! In this month's edition, we take a look at the power (and unfortunate illusion) of ethics teams, AI and human alignment, and more. 

Don't forget: At the bottom of this digest, you have the opportunity to ask a data scientist any question on your mind and we'll feature the response in an upcoming edition.

As always, thank you for your commitment to adopting and building more responsible AI systems!

Cal Al-Dhubaib, CEO

What Do We Mean by Trusted AI?

Trusted AI is the discipline of designing AI-powered solutions that maximize the value of humans while being more fair, transparent, and privacy-preserving.

Learn More

The Latest in Trusted AI

Are AI Ethics Teams Doomed To Be a Facade?

What it's about: In the age of trusted AI, we've seen countless tech giants create alleged AI "ethics" teams that crumble when confronted with serious AI issues. In this article, VentureBeat interviews several AI ethics women pioneers to explore how these teams could succeed when done right.

Why it matters: Before you deploy AI solutions at your company it is critical that you have a person or team of people responsible for evaluating and safeguarding against any unintended ethical consequences. Take some of these powerful tips to create an effective ethics team within your own organization.

Read More

OpenAI Solves For AI + Human Alignment Challenge

What it's about: The challenge of ensuring machine learning models act in accordance with human intentions is known as the "alignment problem." In this article, OpenAI shares how they tested scalable alignment techniques through book summary generation. 

Why it matters: Trusted AI solutions must be made for humans and involve humans. But how can humans oversee an AI model's work when it involves massive amounts of data (e.g. a full novel)? The short answer: We can't unless scalable techniques, like the ones OpenAI is working on, are created.   

View the Experiment

Google's Commitment to Responsible AI in Search

What it's about: For decades, Google has used innovative language models to power Search. In this article, Google shares some ways they're paying attention to the risks and challenges associated with AI in Search. 

Why it matters: We've seen tech giants, including Google, come under fire quite a few times for their questionable AI models. But it's evident in this article that the company is trying to be proactive with a transparent, trusted AI message. Does your company have a similar message available to customers and other stakeholders?

Read More

Connect With Trusted AI Experts

Meet Joy Buolamwini, Founder and Poet of Code

Joy Buolamwini is a poet of code who uses art and research to illuminate the social implications of artificial intelligence. She founded the Algorithmic Justice League to create a world with more equitable and accountable technology. Her TED Featured Talk on algorithmic bias has over 1 million views. Her MIT thesis methodology uncovered large racial and gender bias in AI services from companies like Microsoft, IBM, and Amazon.

Her research has been covered in over 40 countries, and as a renowned international speaker she has championed the need for algorithmic justice at the World Economic Forum and the United Nations. She serves on the Global Tech Panel convened by the vice president of European Commission to advise world leaders and technology executives on ways to reduce the harms of A.I. Connect with Joy.

Your Trusted AI Questions, Answered

Who's responsible for AI's mistakes?

While we would all love to blame the "Malevolent AI Robot", ultimately AI’s mistakes are those of the humans behind the solution.

From the initial stages of AI development when the ultimate product is just an idea, all the way through fully functional operation, there are conversations to be had and questions to be asked that are essential to mitigate risk.

Who is this project going to be affecting? What decisions are going to be made? What is the potential for disparate impact? What are some unforeseen consequences? How are we measuring and tracking this information along the way?

The more complicated the AI solution, the more important it is to ask these questions early and often and design the system with as much interpretability and transparency as possible.

This is not to say that humans, or AI, are ever going to be perfect. However, with proper discussion, investigation, and transparency, there will be safeguards in place to prevent or catch mistakes early on.

- Hannah Arnson, PhD // Director of Data Science