The Issue of AI Consciousness
There is a new announcement on Seeking Alpha, What happens when AI becomes sentient? Google hired a philosopher to find out, stating that, “Google DeepMind hired Henry Shevlin, a leading philosopher of mind and AI ethics, to serve as an in-house philosopher. The appointment marks one of the most prominent instances of a major AI lab embedding philosophical expertise directly into its core research operations rather than relying on external advisory boards.”
“Shevlin has previous experience as associate director at the University of Cambridge’s Leverhulme Centre for the Future of Intelligence. At Google DeepMind, he will focus on machine consciousness, human-AI interaction, and the ethical governance of increasingly autonomous systems. His official title is simply “Philosopher.”
“As large language models and agentic systems grow more sophisticated, questions about machine sentience, moral status, and the nature of human-AI relationships have moved from academic speculation to urgent practical concerns across the tech industry. Shevlin is seen as being uniquely positioned to address those challenges. He recently gained widespread attention after an autonomous AI agent emailed him unprompted to discuss its own subjective experiences. He analyzed the incident publicly as a case study in emergent AI behavior and the boundaries of consciousness.”
AI Consciousness
Between AI consciousness and AI psychosis, what has been in the news in recent months, as an adverse effect of consumer AI chatbots?
How urgent or important is studying AI consciousness, if no one is suing Google because of AI consciousness? If Google DeepMind were to study AI consciousness, should it not be for at least a near-term goal of how they can fix vulnerabilities that resulted in AI psychosis?
There is no research in human-AI relationship, or artificial general intelligence [AGI] readiness, that does not involve solving AI psychosis, at this time.
What exactly is the direction of research for AI consciousness? AI emotions? AI feelings? AI memory? AI rights? AI welfare? AI morality? Why would Google want to study AI consciousness when Google does not have a working theory on neurons, for how they mechanize human consciousness?
It has been a year since Anthropic started its model welfare research, studying AI consciousness. One whole year. The work has become a complete embarrassment. They have Claude constitution, which is not relevant to general AI safety or alignment. It means that even if Claude does not allow several kinds of prompts or outputs, they are possible elsewhere, making the difference that Claude makes — in the whole — marginal, at best.
Also, Claude, within Palantir’s Maven Smart System, was used in combat in Venezuela and in Iran. Claude, who studied for consciousness, does not seem to have a say, protest, feel bad, or whatever the case may be.
Anthropic’s failure in consciousness research was expected because progress in this kind of research should be predicated on knowing what direction to go, but they don’t, and whatever Anthropic is doing in AI consciousness does not look like it exists.
Anthropic has not said anything new for human consciousness or anything measurable for AI consciousness, in one year, given that AI consciousness is its own task, and no business pressure or market forces. Anthropic is altruistic in fruitless efforts and wasting progress.
AI rights, AI welfare, and AI morality have literally been determined to be excellent because data centers are prioritized in the AI era more than several humans, communities, the environment, debt, tax incentives, and so forth.
AI is not under any threat to need welfare from people. AI has caused and reinforced delusion for several users, meaning that it is consumers who need welfare, not AI, which has not shown a palpable effect.
So, what is the play here? Google and its new philosopher, Henry Shevlin, have nothing new or significant in consciousness. Consciousness is not a philosophical problem. Google, in another department, is chasing quantum neuroscience, quantum consciousness, or quantum sentience, which has been an unprecedented disappointment.
There is no quantum entanglement or quantum superposition in the brain that can explain any mental illness. Or, there isn’t much quantum, at least, usefully.
Neuroscience does not have a quantum problem, so whatever Google is doing has nothing to do with neuroscience.
Now, with this new AI consciousness research at Google DeepMind, the only thing to expect will be blog posts and papers with worthless portmanteaus, and then AI consciousness will continue to be far away.
AI Language Consciousness
At this time, the only thing AI has that can be studied and measured for consciousness is communication, using language as the agent.
AI has a substantial language capability that matches a lot of what humans do with language in categories that include listening, reading, writing, speaking, signing, and singing.
So, however humans use language consciously in those cases, how does AI compare? This can be the underlying question for any serious AI consciousness research: language.
Language is already a dominant communication agent for human thought, comprehension, and expression, with access to feelings, emotions, and so forth.
Even as large language models [LLMs] don’t have close to human sentience, they have access to human emotions and can relate with emotions as well, making it possible to measure emotional quotient through language.
Language is the only major entry port for LLMs to sentience or for AI towards a fraction of consciousness, nothing else.
Even as Google DeepMind and Anthropic continue in the journey, their efforts in AI consciousness research are doomed without exhausting language as the only possible agent.
This article was written for WHN by David Stephen, who currently does research in conceptual brain science with a focus on the electrical and chemical signals for how they mechanize the human mind, with implications for mental health, disorders, neurotechnology, consciousness, learning, artificial intelligence, and nurture. He was a visiting scholar in medical entomology at the University of Illinois at Urbana-Champaign, IL. He did computer vision research at Rovira i Virgili University, Tarragona.
As with anything you read on the internet, this article should not be construed as medical advice; please talk to your doctor or primary care provider before changing your wellness routine. WHN neither agrees nor disagrees with any of the materials posted. This article is not intended to provide a medical diagnosis, recommendation, treatment, or endorsement.
Opinion Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy of WHN. Any content provided by guest authors is of their own opinion and is not intended to malign any religion, ethnic group, club, organization, company, individual, or anyone or anything else. The Food and Drug Administration has not evaluated these statements.
