Confessions of an AI Brain

Confessions of an AI Brain, Second Edition invites readers to not only to understand how AI works, but to imagine, question, and actively shape the kind of AI‑enabled future they want to build. By framing AI as a narrator, the book ventures inside the mind of MIRANDA, a fictional yet technically grounded AI brain that narrates its own coming‑of‑age in a world saturated with algorithms, data, and digital infrastructure. Blending clear explanations with an engaging first-person voice, the book demystifies how modern AI systems are built, trained, deployed, and governed―without assuming any mathematical or programming background. Along the way, MIRANDA reflects on topics such as learning from data, bias and fairness, neurosymbolic reasoning, agentic AI, foundation models, digital twins, and the infrastructure that makes intelligent services possible.

Rather than treating AI as a black box or distant threat, the narrative frames it as a partner in a shared human–machine ecosystem, exploring how artificial and biological minds co‑evolve at home, in industry, in cities, and in our creative lives. Through concrete examples ― from baby steps in safe sandboxes and “data diets” to multi‑agent systems, climate‑aware optimization, and AI‑assisted art, music, and writing ― the book shows how AI can be designed to support sustainability, inclusiveness, and human flourishing. Readers gain an intuitive grasp of core concepts like machine learning styles, attention and transformers, trustworthy and explainable AI, and the role of regulation and ethics, all framed through MIRANDA’s candid “confessions” about its capabilities and limitations.

Aimed at curious general readers as well as professionals and students who want a concept-focused overview, Confessions of an AI Brain bridges technical insight and societal reflection in an accessible, story‑driven format. It is also a useful companion for engineers, business leaders, policymakers, creatives, and anyone adopting AI in their work and want to ask better questions about data quality, safety, bias, responsibility, and long‑term impact.

Elena Fersman is a Vice President, Head of AI Innovation and Incubation and Head of Ericsson Silicon Valley Site at Ericsson. Elena is a docent in Cyber-Physical Systems specialized in Automation at the Royal Institute of Technology in Stockholm, holds a PhD in Computer Science from Uppsala University, and did a postdoc at the University Paris-Saclay. At Ericsson, she had various positions ranging from product management to research leadership. Elena is a fellow of the Royal Swedish Academy of Engineering Sciences. Elena has co-authored over fifty patent families.

Paul Pettersson is Vice-President and Professor at Mälardalen University, Sweden. He has a PhD in Computer Systems and has been appointed Docent in Computer Science at Uppsala University, Sweden. Prof. Pettersson has co-authored over 100 scientific papers, articles, and book chapters in Computer Science. During his career, he has held various leadership positions in academia, including six years as President (Vice-Chancellor) of Mälardalen University.

Athanasios Karapantelakis is a research engineer with seventeed years of experience in the telecommunications and software industries. His research focuses on automated planning and reasoning and their integration with machine learning. He has held diverse engineering roles spanning software design, cloud technologies, and AI research. Athanasios holds a PhD from KTH Royal Institute of Technology, Stockholm, Sweden.

Christina Chaccour is a Senior Strategy and Product Manager at Ericsson working on AI for networks and emerging wireless technologies. Her background spans telecommunications systems, AI, and microelectronics, with roles across research, technical pre-sales, network architecture, and business strategy. She holds a PhD in Electrical Engineering from Virginia Tech, represents Ericsson in key industry forums and on editorial boards, and is a co-founder of an award-winning technology startup.