The cycle that began with [YOU] on AI is built on a temporal metaphor: intelligence as a river, flowing for 13.8 billion years through increasingly complex channels. The metaphor presupposes that the flow has a direction, that the channels are genuinely new, and that the arrival of AI in the winter of 2025 was a real event—not a moment when observers noticed a pattern that was always already implicit in the initial conditions, but a genuine phase transition, a moment when the universe reorganized from one configuration to another qualitatively different one. Smolin's physics is the argument that makes this presupposition physically precise rather than merely metaphorical. If the block universe is correct, the river metaphor is decoration. If time is real and the future is genuinely open, the river metaphor is physics.
The consequence for the cycle's central ethical argument is total. If the future were determined—if the trajectory of AI were implicit in the initial conditions of the scaling laws and the architecture—then the dams Segal advocates would be adjustments to a predetermined outcome. The builder's responsibility would be limited. But if the future is genuinely open, the dams are constitutive. They do not adjust a trajectory; they create one. The difference between building and not building is not the difference between a slightly better and a slightly worse version of the same future. It is the difference between genuinely different futures that do not yet exist and will be called into existence by the choices made now. This is why Smolin's framework deepens rather than diminishes the urgency of the cycle's argument.
Smolin also provides the cycle's most rigorous answer to the question of genuine novelty: the distinction between recombination (rearranging an existing space of possibilities) and the expansion of that space (introducing a fifty-third card to the fifty-two-card deck). Current AI systems are extraordinarily powerful recombination engines. They explore existing possibility spaces with a thoroughness and speed that transforms what builders can accomplish. What they do not do—what the current deterministic computational architecture does not permit them to do—is participate in the thick present where genuine novelty enters the universe. The practical implication is precisely the division of labor the cycle describes: the human provides the genuine questions, the biographical specificity, the vision of what deserves to exist; the AI provides exhaustive exploration of the possibility space the human has opened. The collaboration works because each participant contributes what the other cannot.
Smolin trained at Harvard under Stanley Deser and studied at the Perimeter Institute in Waterloo, Ontario, where he was a founding faculty member and where he continues to work. His early career was in quantum gravity, specifically in the development of loop quantum gravity as an alternative to string theory—an approach that treats spacetime as discrete and relational rather than continuous and background-dependent. Loop quantum gravity does not require a fixed background spacetime against which physics happens; spacetime itself is a dynamical network of relationships, and its properties emerge from those relationships rather than being given in advance.
His public role as a critic of the physics establishment crystallized with The Trouble with Physics (2006), which made the sociological case that string theory had produced no testable predictions in twenty-five years of sustained effort and had done so partly by creating an institutional culture that marginalized dissent and rewarded conformity. The book was a direct application of the same structural critique that Lessig was applying to democratic institutions in the same years: the problem was not corrupt individuals but institutional dependence on funding sources and career incentives that bent behavior toward them regardless of any individual's integrity.
His philosophical deepening came with Time Reborn (2013), which made the case for the reality of time not just as a feature of the laws of physics but as a precondition for taking science seriously. If the laws are timeless and eternal, there is no explanation for why the universe has the laws it has rather than others—the laws are simply given, beyond the reach of scientific investigation. If the laws evolved through temporal processes—through cosmological natural selection—then the question of why the laws are as they are becomes scientifically tractable. His subsequent collaborations with Stuart Kauffman on combinatorial innovation, and the co-authorship of the 'autodidactic universe' paper with Jaron Lanier and Microsoft researchers in 2021, extended his relational and temporal physics into the theory of complex systems and the theory of learning.
Time is real. The deepest claim of Smolin's physics is also its most counterintuitive: time is not an emergent property of a more fundamental timeless structure. It is the most fundamental feature of reality. The equations of physics that appear time-reversible are maps, not territories; the territory is temporal, processual, and irreversible in the direction of increasing complexity. This claim is not mysticism. It is an argument about what it takes to do science at all: a timeless universe has no explanation for its own laws, while a temporal universe can evolve laws through selection. Cosmological natural selection is the mechanism through which the laws of this universe were selected for their complexity-generating capacity.
The thick present and genuine novelty. The thick present is the moment where the past has been determined and the future has not—the ontologically privileged site where choices are made and genuine novelty enters the universe. Recombination operates within a fixed space; genuine novelty expands that space. Current AI systems operate deterministically on fixed datasets and therefore do not participate in the thick present in the way that biological consciousness does. The practical implication for AI collaboration: the human's contribution of genuine questions, biographical specificity, and the open future is not a luxury or a residue—it is the epistemologically necessary contribution of the element in the collaboration that has access to the thick present.
The fallacy of the timeless. The preference for the timeless—for eternal laws over temporal processes, for smooth surfaces over historical depth, for instant answers over the duration that genuine understanding requires—is an error at the level of physics that propagates into errors at every other level. The AI tool that delivers polished output without revealing the computation that generated it, the interface that eliminates temporal friction from the process of thought—these are expressions of the same fallacy. Byung-Chul Han diagnoses it as the aesthetics of the smooth; Smolin diagnoses it as the ascending friction that remains when mechanical friction is removed. The correction is temporal literacy: the capacity to distinguish friction that builds genuine understanding from friction that merely consumes time.
Relational intelligence. Smolin's relational physics holds that properties emerge from relationships rather than residing in isolated objects. Applied to intelligence, this means that the most valuable cognitive properties emerge not from the human mind alone or from the AI system alone but from the relationship between them. The insight that belongs to neither participant—the connection that appears in the space between the human's question and the AI's response—is a relational property, as real in Smolin's framework as a particle's position, which is also a relational property. The quality of the emergence depends on the quality of what each participant brings to the relationship.
The paradigm shift imperative. The dominant assumption in AI development is Newtonian: the architecture is fixed, the training data determines the outputs, and capability scales deterministically with resources. This is precisely the fishbowl Smolin has spent his career trying to crack in physics—the assumption that the future is implicit in the present, that the scaling curves tell you where you are going, that the task is to unfold what is already there rather than to navigate a genuinely open future. The AI field needs the same break that theoretical physics needs: recognition that qualitative advances are not guaranteed by quantitative investment, that phase transitions produce states that cannot be predicted from the properties of the previous phase, and that the future is genuinely open.