The controversial idea that consciousness has a quantum origin is facing a new challenge, after a mathematical analysis appears to show that a long-known law of physics makes it impossible.
What physical process makes us conscious, and whether we could observe and analyse it, is one of the great open questions of modern science. In the 1990s, mathematician Roger Penrose and anesthesiologist Stuart Hameroff proposed that classical physics can鈥檛 explain consciousness alone and so it must arise in the quantum realm. The idea was received with scepticism and remains controversial, but attempts to bring quantumness into explanations of consciousness have persisted since.
Now, at Chapman University in California and her colleagues have examined the idea of purely quantum agency, which is a prerequisite for consciousness, and found it doesn鈥檛 work.
Advertisement
Consciousness is notoriously hard to define, but Adlam and team formulated a minimal definition of agency using clear physical requirements. Quantum theories of consciousness, even if they focus on specific mechanisms in the brain, like in Penrose and Hameroff鈥檚 proposal, ought to be compatible with this definition if they are to be plausible. Specifically, the researchers focused on a model of an agent that observes the world, creates a model of it, then uses that model to evaluate outcomes of different possible actions before deciding on performing one. 鈥淭here will be a step in terms of taking information out of the environment that you鈥檙e going to respond to, and then some kind of [information] copying steps, so that you can consider different possible options, and then at the end, a comparison step where you look at the outputs of each of your models, decide which is the best and use that to guide your actions,鈥� says Adlam.
But if the agent is quantum 鈥� so must behave according to the laws of quantum mechanics 鈥� the copying step is impossible. This is because of the 鈥渘o-cloning theorem鈥�, which says that quantum states that describe all information about a system can鈥檛 be copied or duplicated.
Adlam says that an agent may try to deal with this by creating imperfect copies of information it needs to deliberate instead, but other problems then arise. For example, extracting quantum information from the world changes it, so the agent鈥檚 world model actually reflects a slightly different world. While some level of imperfection is always present in how we model the world, these faults would compound.
The comparison step also suffers from quantum weirdness. Adlam says that this is because the purely quantum agent could only use quantum information processing, which makes available different processes than we are used to. 鈥淎 realistic quantum operation is going to end up doing some kind of funny superposition of all of the possible options, weighted towards the better ones. You don鈥檛 actually implement the best option, you do this weird combination of all of them,鈥� says Adlam.
These conclusions challenge quantum theories of agency, free will and consciousness, as well as impose technological constraints on any agents that quantum computers may be able to simulate in the future, said team member , also at Chapman University, in a presentation at the聽聽conference in Irvine, California, on 17 June. 鈥淭hese results help us understand how agency and human-like perception can emerge in an otherwise quantum mechanical universe,鈥� he said.
In fact, the problems with a purely quantum agent that the researchers identified make it harder to justify an essentially quantum mechanism for both agency and consciousness. It is much more likely that these phenomena arise from, or are at least dominated by, classical mechanisms, says Adlam.聽鈥淭here could be some circumstances where perhaps doing a quantum calculation gives you a speed-up, but certainly it makes the whole process of agency look like it is basically founded in the classical regime,鈥� she says.
鈥淨uantum physics is the best physical theory that we have, so it makes sense for us to ask, what does agency look like from the perspective of one of our best physical theories?鈥� says at the University of Queensland in Australia. In her view, the researchers鈥� model of agency is informed by a specific interpretation of what is essentially real in the quantum world. If quantum states are seen not as idealised building blocks of reality, but rather objects that reflect our knowledge of the world, then a quantum model of agency could be less constraining, she says.
For example, an agent could be modelled as continuously interacting with an environment and dynamically converging on a decision instead of deliberating between world models, which would allow for a combination of classical and quantum information processing.聽 Shrapnel says that all existing theories of quantum consciousness aren鈥檛 as strictly quantum as the model of agency in the new study, so they may not be fully off the table just yet.
While consciousness remains very difficult to study directly, Adlam and Shrapnel both say that future experiments with quantum computers may shed more light on the mixing of quantum and classical information processing that may pertain to this discussion.
