We must explain! We will explain!
Published:
CWe must explain! We will explain!
What if AI could be explained — and what if it must be?
Join Beta and Bit, two fearless adventurers in the world of artificial intelligence, as they traverse iconic research hubs like Singapore, Berlin, Pisa, London, and Warsaw to confront one of the most pressing challenges in modern AI: the need for explainability. With style, wit, and rigor, this comic dives into historical, philosophical, and scientific debates—from Gödel’s theorems to Hinton’s neural net critiques—making a compelling case for post-hoc interpretability techniques in AI systems.
Science, ethics, and storytelling collide in this visually stunning journey.
From uncovering biases in recidivism risk scores to learning strategy from AlphaZero, Beta and Bit collect evidence that explanations aren’t just possible—they’re essential. We Must Explain! is more than an accessible introduction to eXplainable AI (XAI); it’s a manifesto for transparency and accountability in machine learning, packaged in a brilliant, illustrated adventure that both informs and inspires.
