We Cannot Transfer Your Consciousness, Unfortunately
Written on
The Fallacy of Mind Uploading
Imagine a future where, as you take your final breath, a preservative is injected into your brain to maintain every one of the intricate connections among your approximately 86 billion neurons. This procedure would allow us to create a detailed map of your brain's connections, known as your “connectome,” and upload this data to a computer, essentially allowing you to live again.
This is the tantalizing promise of mind uploading. It's based on the belief that the connections between neurons hold the key to your memories and identity. Various organizations are striving to make this a reality by discovering the best methods to preserve these intricate wiring diagrams. Enthusiastic media reports, academic papers, and controversial advocates promote the idea that we could restore you from your neural connections. As Sebastian Seung famously stated, "You are your connectome."
However, this assertion is fundamentally flawed. The concept of mind uploading from a preserved brain is unattainable—not just due to our lack of advanced technology, but also because it is impossible to recreate a person solely from their neural wiring.
Understanding the Disconnect
The reasons for this are straightforward. A connectome represents a static snapshot of the brain's wiring at the precise moment it was recorded, while the brain itself is a dynamic entity. It is constantly changing, with billions of neurons communicating and reshaping their connections, as well as the complex molecular processes occurring within each neuron that dictate its form and function.
Focusing on one aspect of this dynamic nature can illustrate the point effectively. Neurons communicate by sending electrical pulses, or "spikes," along their axons to other neurons. This activity is foundational to all our functions. For instance, spikes from motor neurons in the spinal cord cause muscle contractions, resulting in movements like flexing fingers or extending arms. Additionally, spikes facilitate sensory perception and decision-making.
However, knowing the wiring of neurons gives us no insight into the spikes they transmit. It only informs us of the pathways available for communication at the moment of preservation, raising a critical question: why would anyone want to be reconstituted at the moment before death, which was likely not their best time?
The Timing and Frequency of Neuronal Spikes
The timing and frequency with which a neuron sends spikes depend on how it reacts to incoming signals. While understanding the type of neuron can provide a general idea of its responses, it only offers qualitative insights. For instance, a pyramidal neuron in the cortex behaves differently than a fast-spiking interneuron, yet we cannot ascertain the specific behavior of an individual neuron just based on its type.
The precise dynamics of a single neuron depend on its unique shape, size, the location of its inputs, and the specific ion channels it expresses. The connectome, however, reveals none of these vital characteristics. We cannot predict how a single neuron will respond to incoming signals based solely on its wiring. Furthermore, when spikes arrive at the synapse connecting two neurons, the impact of that spike is also not discernible from the connectome.
Two critical questions arise that the connectome cannot address:
- Connection Strength: The strength of the connection between neurons affects how influential a spike from one neuron is on the behavior of another. While there are some correlations between synaptic features and strength, such as spine size or neurotransmitter receptor count, these aspects cannot be deduced from wiring alone. Accurately determining the strength of a synapse requires extensive experimental investigation, which is unfeasible for the roughly 1 trillion synapses in the brain.
- Connection Reliability: Some synaptic connections are consistently reliable, while others may fail frequently. For example, connections in the hippocampus can fail up to 90% of the time. This variability in reliability is also invisible to the connectome. Understanding the frequency of synaptic failures is crucial for accurate simulation of brain function.
Additionally, both connection strength and reliability are not static; neurons can grow or shrink their dendrites, inputs change dynamically, and the expression of ion channels fluctuates. This ever-evolving nature of the brain defines who you are—your thoughts, memories, and actions.
The Limitations of a Static Model
Attempting to recreate your brain based solely on neural wiring is akin to trying to reconstruct your entire social life by merely analyzing your current contacts. It overlooks the complexities of interactions—how often you engage with someone, the quality of those interactions, and the historical context of each relationship.
Even if we had complete knowledge of the structure, size, and ion channel composition of every one of your neurons, along with the strength and reliability of every synapse, it would still fall short. The timing and frequency of neuronal spikes primarily depend on the spikes received from other neurons. This creates an infinite loop of interactions that cannot be simulated without understanding the entirety of your neural development, starting from the moment your first spikes occurred.
While there are compelling scientific reasons to develop connectomes—such as understanding neuronal wiring principles, exploring brain variability among individuals, and identifying wiring changes linked to mental health disorders—connectomes remain static entities, whereas you are not. Therefore, it is with regret that I must inform you that uploading your consciousness to a computer is simply not feasible; your identity cannot be confined to the wiring of your neurons. You are not your connectome.
Mark Humphries is a researcher in computational neuroscience at the University of Nottingham, UK, and author of The Spike: An Epic Journey Through the Brain in 2.1 Seconds (Princeton University Press).
Twitter: @markdhumphries