A mind upload captures a brain at a specific moment in time. But which moment, and how does that moment relate to what came before and what would have come after? The question of lifespan matters for brain emulation because the functional organization of a brain at age 35 is not identical to its organization at age 65, and the differences are not random degradation. They follow structured developmental trajectories that have now been mapped with unprecedented scope.

A Nature 2026 paper by Taylor and colleagues presents the first comprehensive atlas of human functional brain organization across 3,500 subjects spanning birth to 100 years of age. The atlas documents how three fundamental organizational axes of the brain emerge in early development, mature through childhood and adulthood, and decline with aging. The research identifies core functional patterns that remain stable across the lifespan and maps the systematic deviations that characterize different life stages.

For whole brain emulation, this atlas serves a function analogous to what structural connectomics provides for anatomical mapping: a reference framework against which any claimed emulation can be compared. An emulated brain should not just preserve the connectivity and firing patterns measured at the time of scanning. It should preserve a brain that sits coherently within the lifespan trajectory its owner was following, at the specific position along that trajectory they had reached.

Three Organizational Axes

The atlas characterizes functional brain organization in terms of principal axes that capture the dominant patterns of correlated activity across brain regions. These are not anatomical regions but functional gradients, continuous dimensions along which brain regions vary in their connectivity profiles and behavioral associations.

The first axis, the one with the longest developmental trajectory, separates the default mode network from sensorimotor systems. Default mode activity increases during internally directed thought, memory retrieval, and social cognition. Sensorimotor regions respond to external inputs and motor outputs. These systems are anticorrelated in adults, but this anticorrelation is not present in neonates. It emerges gradually through childhood and reaches adult form in adolescence.

The second axis distinguishes frontoparietal control networks from limbic and paralimbic regions. This gradient is associated with cognitive control, the ability to regulate behavior flexibly in response to changing goals, and it undergoes substantial refinement through early adulthood. Individuals show the greatest individual variation along this axis in their twenties, a period associated with identity consolidation and the peak of creative and entrepreneurial achievement in population statistics.

The third axis, most prominent in aging data, separates heteromodal association cortex from unimodal sensory regions. Its decline in later life correlates with the cognitive changes that characterize normal aging and, when accelerated beyond age-typical norms, with the early stages of neurodegenerative disease.

What the Atlas Contributes to Emulation Science

The Meta TRIBE v2 brain digital twin demonstrated that functional brain responses can be predicted from individual-level data with 70,000-voxel resolution. That work established that individual brains have distinctive functional signatures that can be modeled. The Taylor 2026 atlas provides the lifespan context that individual-level models need: any given brain’s functional signature exists against the backdrop of what functional organization typically looks like at that age, in that developmental phase.

The Virtual Brain Twins clinical framework uses patient-specific brain models for treatment optimization in epilepsy and multiple sclerosis. These models are calibrated to individual structural and functional data but do not currently incorporate lifespan position as a modeling parameter. The atlas provides the normative reference that would allow such models to distinguish disease-related deviations from age-typical variation.

For brain emulation more specifically, the lifespan atlas raises a question about what it would mean for an emulation to be faithful. An emulation of a 70-year-old brain should not function like a 30-year-old brain, even if the 70-year-old’s structural connectome is intact. The functional organization changes with age independently of structural connectivity. Preserving the structural connectome without preserving the age-appropriate functional state would produce an emulation that behaves differently from its biological original.

The Stable Core

Not everything changes with age. One of the more striking findings from the atlas is the identification of a stable core of functional organization that persists from early childhood through late life. This core involves the spatial relationships between primary sensory and motor regions and constitutes the scaffolding on which more flexible functional networks are built and later dismantled.

The existence of a stable core has implications for what must be preserved in any emulation. Some functional organization is essentially invariant across the lifespan and would therefore be correctly captured by any snapshot of brain connectivity taken at any adult age. The variable components are those that change systematically with development, learning, and aging.

The atlas allows decomposition of any individual brain’s functional organization into its stable and variable components. A brain emulation that correctly captures the stable core but misrepresents the variable components would produce behavior appropriate for a different life stage than the original. Whether this constitutes a failed emulation depends on what degree of fidelity is required, but it defines a specific, measurable error type.

Aging as a Biological Process, Not Just Degradation

A recurring assumption in mind uploading discourse is that capturing a brain at a young age, before cognitive decline, would be optimal. The lifespan atlas complicates this framing. The functional organization of a 35-year-old brain and a 65-year-old brain are not simply different quantities of the same quality. They are different organizational states, each reflecting different balances between cognitive systems shaped by decades of different experience.

The reduction in default mode to sensorimotor anticorrelation that occurs in older adults, for example, may not represent degradation in any straightforward sense. Research reviewed in the atlas suggests it reflects a shift in how cognitive resources are allocated across task states, a shift that trades processing speed for integrative processing of complex information. An older brain that appears to have lost the sharp network segregation of youth may instead have achieved a different kind of functional integration.

This matters for the question of what a person would be if their brain were emulated at age 70 rather than 35. The lifespan atlas suggests the answer is not simply a degraded version of their 35-year-old self but a distinct functional state shaped by everything that happened between those ages. The 4E cognition challenge to mind uploading makes a related point from a philosophical angle: cognition is not a static software state but an ongoing process embedded in history and environment. The atlas provides empirical backing for the temporal dimension of that argument.

Caveats and Limitations

The atlas is built from cross-sectional data: 3,500 different individuals at different ages rather than longitudinal tracking of the same individuals across decades. Cross-sectional samples are the only feasible approach for lifespan data at this scale, but they introduce cohort effects. People currently aged 80 grew up in different environmental and educational conditions than people currently aged 30. Some apparent age effects in the atlas may reflect generational differences rather than biological aging trajectories.

The atlas captures functional organization as measured by resting-state fMRI, which reflects slow hemodynamic responses to neural activity. Faster neural dynamics, including the millisecond-scale spike timing that determines circuit-level computations, are not captured by this modality. The physics wall facing non-destructive brain scanning applies here: fMRI provides population-level functional organization data, not the individual-synapse precision required for true brain emulation.

What the atlas provides is a normative framework: the population-level map against which individual brain functional organization can be positioned. That is valuable for validation even if it is not sufficient for emulation.

Technology Readiness Level: Not applicable. The atlas is a scientific reference resource. Its value for emulation validation is indirect but substantial.

Official Sources

• Nature (April 2026): Taylor et al., “Charting the Human Brain’s Lifelong Functional Organization”
• Nature Atlas coverage: https://www.nature.com/articles/d41586-026-00975-1
• Related Virtual Brain Twins: DOI 10.1109/RBME.2023.3346589
• Related Meta TRIBE v2 (March 2026): Meta Research blog