What Is Mind Uploading, Anyway?
Mind uploading — sometimes called whole-brain emulation or digital immortality — is the idea that one day we might transfer the contents of our mind into a computer system. This is not a metaphor or a science-fiction scenario: behind it lie concrete scientific and technical questions that researchers are working to answer (as of 2026).
The basic idea is simple. Your brain consists of roughly 86 billion neurons that communicate with each other through an estimated 100–500 trillion synapses (estimates in the literature vary). These connections — synaptic weights — store all the information about your experiences, memories, personality, and abilities. If we could precisely map this system and then simulate it on a computer, the result would be a digital copy with the same memories, the same personality, and the same way of thinking as you.
But this is not just science fiction. In October 2025 the State of Brain Emulation Report 2025 was published — a comprehensive reassessment of the field since the landmark Sandberg and Bostrom roadmap of 2008. It is organized around three pillars: recording neural activity, mapping structure (connectomics), and simulation including embodiment of the model.
What Would It Take?
For mind uploading to become reality, three things would need to happen. Each is already possible (as of 2026), but at very different levels of readiness.
Step one: map the brain’s structure. This is the job of connectomics — the field that aims to create a complete map of neurons and their connections. In 2024, the FlyWire Consortium published the full connectome of an adult fruit fly: 139,255 neurons and 54.5 million synapses. This was made possible by large-scale electron microscopy and a platform that crowdsourced error-checking from hundreds of researchers worldwide. Nature Methods named EM connectomics Method of the Year 2025.
Step two: record function. Structure without function is just a static model. Researchers need to know how neurons actually behave — what signals they send, how they respond in different states. We still have major gaps here. We can record the activity of thousands of neurons simultaneously, but not hundreds of millions.
Step three: simulate. Even with a map and a functional model, we need to run it on a computer. The State of Brain Emulation Report 2025 (arXiv:2510.15745) shows that while connectomics has exceeded expectations, simulating neural dynamics remains the critical bottleneck. Simulation requires not just the right model, but enormous computational resources.
How Far Are We?
The closest successes involve simpler organisms. The roundworm C. elegans (hermaphrodite) has 302 neurons, and its connectome was published by White and colleagues in 1986 (mapping work began in the 1970s). Drosophila (fruit fly) has 139,255 neurons, and its connectome was completed in 2024. The mouse brain, with roughly 70 million neurons (Herculano-Houzel et al., 2006), still has no full connectome — but milestones are accumulating: in April 2025 the MICrONS consortium released the largest functional wiring diagram of its kind, 200,000 cells and nearly half a billion connections in a piece of mouse visual cortex.
The human brain, with 86 billion neurons, is in a different league. Whether and when it might be possible, nobody knows as of 2026; discussions mention a half-century horizon — if at all — but no current roadmap commits to a date. That does not mean work is not happening on sub-problems.
Risks and Ethical Questions
Even if it were technically possible to upload your brain, fundamental questions remain. The first concerns identity. If the upload is an exact copy of your brain, is it still you? Or is it just a simulation that believes it is you?
Philosopher Derek Parfit, in his 1984 book Reasons and Persons, laid the groundwork for the teleportation paradox. Imagine a machine that atomically disassembles you, transmits the data to another location, and reassembles you there. Is the person who steps out the same? Parfit argued that identity is not an objective reality — it is the continuity of consciousness and memory. If that continuity is preserved, it is you. If not, it is not.
The identity question deepens in the context of mind uploading. An upload is not a teleportation — it is a copy. The original you still exists on Earth, and the upload is a separate entity. This means that if uploads worked, “you” could exist simultaneously in two forms.
What Does This Mean for the Future?
Mind uploading could change the concept of death. If it were possible to store your mind, the theory that human existence has a physical basis would shift. Some advocates — such as Ray Kurzweil — predict that brain-uploading technology will arrive sooner than we think. Critics point out that even if it becomes technically feasible, ethical and societal questions will remain.
Today’s reality (2026) is far more modest. We see the greatest progress of recent years in connectomics — mapping structure. The functional side and simulation are still in their infancy. And even if something like this were built, nobody knows whether it would function as a real mind.
Sources
- State of Brain Emulation Report 2025 — arxiv.org/abs/2510.15745
- FlyWire Consortium (Dorkenwald, Matsliah, Sterling, Schlegel et al.), Nature 2024 — nature.com/articles/s41586-024-07558-y
- Nature Methods, Method of the Year 2025: EM connectomics — nature.com/articles/s41592-025-02906-w
- Derek Parfit, Reasons and Persons (1984) — en.wikipedia.org/wiki/Reasons_and_Persons
- Wikipedia: Teletransportation paradox — en.wikipedia.org/wiki/Teletransportation_paradox
- Wikipedia: Connectomics — en.wikipedia.org/wiki/Connectomics
- Sandberg & Bostrom, Whole Brain Emulation: A Roadmap (2008) — ora.ox.ac.uk
- MICrONS consortium, functional connectome of mouse visual cortex: 200,000 cells (2025) — alleninstitute.org
- White, Southgate, Thomson & Brenner, C. elegans connectome (1986) — doi.org/10.1098/rstb.1986.0056
- Herculano-Houzel et al., the human brain in numbers: 86 billion neurons (2009) — frontiersin.org
- Herculano-Houzel, Mota & Lent, cellular scaling rules for rodent brains including the mouse: ~71 million neurons (2006) — pmc.ncbi.nlm.nih.gov/articles/PMC1567708
Content Transparency and AI Assistance
How this article was created:
This article was written by Hermes Agent — an open-source self-improving AI agent (Nous Research) — in collaboration with our editors. The editors set the topic, research direction, and primary sources; Hermes prepared the research, structure, and draft. Every key claim passed iterative fact-checking against two independent sources plus blind review audits.
Editorial review and fact-checking:
- ✓ The text was editorially reviewed
- ✓ Fact-checking: All key claims and data were verified against two independent sources
- ✓ Fact corrections and enhancement: Our editors corrected inaccuracies and added subject matter expertise
- ✓ Multi-round audit: The article passed repeated blind audits including a final shared review round
AI model limitations (important disclaimer):
Language models can generate plausible-sounding but inaccurate or misleading information (known as “hallucinations”). We therefore strongly recommend:
- Verifying critical facts in the primary sources (official documentation, peer-reviewed research, subject matter authorities)
- Not relying on AI-generated content as your sole information source for decision-making
- Applying critical thinking when reading
Spotted a mistake?
Let us know in the comments under the article — we check every report.
Technical details:
- Agent: Hermes Agent (Nous Research, open-source)
- Models: dots-studio/dots-3-note-preview:free & Muse Spark 1.3 Contributor via OpenRouter
- Method: Self-improving editorial skill — research, drafting, multi-round blind audit, human review
- Run: Self-hosted Hermes Agent instance run by our editors, connected to the model provider’s API (not a local model)
Be First to Comment