Why AI should be explainable

One of the biggest challenges of AI is building trustworthy AI-powered systems. Humans and corporations are increasingly uncomfortable with “black box” AI algorithms that offer no insight into how decisions were reached. Our goal is to bring transparency to the models, facilitate compliance, improve model performance and reduce ethical bias.

How we will do it

NukkAI uses the game of Bridge as a testbed for its innovative approaches. Bridge, unlike Chess or Go, has resisted AI so far due to being a multiplayer, incomplete information game whose rules impose some explainability.

We will build models and programs capable of outperforming humans at Bridge, while providing transparency and explanability to their actions. The techniques learned along the way can be applied to real-world domains with great societal benefits.

How we tackle business

We leverage our innovative methods of Explainable AI on several business verticals, combining both “mission driven” and “market driven” approaches. Our approach to business activities revolves around the following three pillars :

• We team up with 3rd parties to identify markets and develop products.
• We create products that can be scaled to other companies within industry, avoiding bespoke, one-off work.
• These initiatives fund NukkAI’s development.