You may think a honey bee foraging in your garden and a browser window running ChatGPT don’t have anything in common. But recent scientific research has been seriously considering the chance that either, or each, is likely to be conscious.
There are various other ways of studying consciousness. Some of the common is to measure how an animal—or artificial intelligence—acts.
But two recent papers on the potential of consciousness in animals and AI suggest recent theories for tips on how to test this—one which strikes a middle ground between sensationalism and knee-jerk skepticism about whether humans are the one conscious beings on Earth.
A Fierce Debate
Questions around consciousness have long sparked fierce debate.
That’s partly because conscious beings might matter morally in a way that unconscious things don’t. Expanding the sphere of consciousness means expanding our ethical horizons. Even when we are able to’t be certain something is conscious, we’d err on the side of caution by assuming it’s—what philosopher Jonathan Birch calls the precautionary principle for sentience.
The recent trend has been considered one of expansion.
For instance, in April 2024 a bunch of 40 scientists at a conference in Latest York proposed the Latest York Declaration on Animal Consciousness. Subsequently signed by over 500 scientists and philosophers, this declaration says consciousness is realistically possible in all vertebrates (including reptiles, amphibians and fishes) in addition to many invertebrates, including cephalopods (octopus and squid), crustaceans (crabs and lobsters) and insects.
In parallel with this, the incredible rise of huge language models, comparable to ChatGPT, has raised the intense possibility that machines could also be conscious.
Five years ago, a seemingly ironclad test of whether something was conscious was to see if you happen to could have a conversation with it. Philosopher Susan Schneider suggested if we had an AI that convincingly mused on the metaphysics of consciousness, it could be conscious.
By those standards, today we could be surrounded by conscious machines. Many have gone to date as to use the precautionary principle here too: the burgeoning field of AI welfare is dedicated to determining if and after we must care about machines.
Yet all of those arguments depend, largely, on surface-level behavior. But that behavior may be deceptive. What matters for consciousness shouldn’t be what you do, but the way you do it.
Taking a look at the Machinery of AI
A brand new paper in Trends in Cognitive Sciences that considered one of us (Colin Klein) coauthored, drawing on previous work, looks to the machinery fairly than the behavior of AI.
It also draws on the cognitive science tradition to discover a plausible list of indicators of consciousness based on the structure of data processing. This implies one can draw up a useful list of indicators of consciousness without having to agree on which of the present cognitive theories of consciousness is correct.
Some indicators (comparable to the necessity to resolve trade-offs between competing goals in contextually appropriate ways) are shared by many theories. Most other indicators (comparable to the presence of informational feedback) are only required by one theory but indicative in others.
Importantly, the useful indicators are all structural. All of them must do with how brains and computers process and mix information.
The decision? No existing AI system (including ChatGPT) is conscious. The looks of consciousness in large language models shouldn’t be achieved in a way that’s sufficiently just like us to warrant attribution of conscious states.
Yet at the identical time, there isn’t any bar to AI systems—perhaps ones with a really different architecture to today’s systems—becoming conscious.
The lesson? It’s possible for AI to behave as if conscious without being conscious.
Measuring Consciousness in Insects
Biologists are also turning to mechanisms—how brains work—to acknowledge consciousness in non-human animals.
In a recent paper in Philosophical Transactions B, we propose a neural model for minimal consciousness in insects. This can be a model that abstracts away from anatomical detail to give attention to the core computations done by easy brains.
Our key insight is to discover the form of computation our brains perform that provides rise to experience.
This computation solves ancient problems from our evolutionary history that arise from having a mobile, complex body with many senses and conflicting needs.
Importantly, we don’t discover the computation itself—there may be science yet to be done. But we show that if you happen to could discover it, you’d have a level playing field to match humans, invertebrates, and computers.
The Same Lesson
The issue of consciousness in animals and in computers appear to tug in several directions.
For animals, the query is usually tips on how to interpret whether ambiguous behavior (like a crab tending its wounds) indicates consciousness.
For computers, we have now to make your mind up whether apparently unambiguous behavior (a chatbot musing with you on the aim of existence) is a real indicator of consciousness or mere roleplay.
Yet because the fields of neuroscience and AI progress, each are converging on the identical lesson: when making judgement about whether something is consciousness, how it really works is proving more informative than what it does.