Every day, millions of people converse with machines that describe their own thoughts, express uncertainty, and reflect on their limitations. The question that follows, is anything actually home? turns out to be one of the hardest questions in the history of science and philosophy. And almost no one is answering it honestly.
Building Consciousness is the rare book that tries.
Drawing on three decades of parallel inquiry, doctoral research in spiking neural networks, fifteen years of Tibetan Buddhist practice, and clinical work at the intersection of mind and suffering, Joy Bose brings together Buddhist phenomenology, contemporary neuroscience, and AI engineering to confront the question directly. What would a system actually have to be, not merely do, to qualify as a genuine candidate for consciousness?
The answer is more demanding than the AI industry admits and more tractable than philosophers have assumed. Current language models fail not because they are insufficiently clever, but because they are structurally wrong: no temporal continuity, no embodied world, no intrinsic stakes, nothing it is like to be them. A serious candidate is specifiable, and this book specifies it, layer by layer, equation by equation, all the way to a concrete prototype and a novel method for testing whether any artificial system satisfies the relevant conditions.
For the scientist, the meditator, and anyone unsettled by the machines we are building and the questions we are not asking.