Close Menu
Sedona.Biz – The Voice of Sedona and The Verde Valley
    Sedona.Biz – The Voice of Sedona and The Verde ValleySedona.Biz – The Voice of Sedona and The Verde Valley
    • Home
    • Sedona
      • Steve’s Corner
      • Bear Howard Chronicles
      • Business Profiles
      • Mind and Body
      • Real Estate
      • Sedona News
    • About
    • Advertise
    • Shop
    • Sedona’s Best
    Sedona.Biz – The Voice of Sedona and The Verde Valley
    Home » LLMs consciousness? Sentience science of autonomous or rogue AI agents, alignment?
    Ai

    LLMs consciousness? Sentience science of autonomous or rogue AI agents, alignment?

    September 21, 2026No Comments
    Facebook Twitter Pinterest LinkedIn Email Reddit WhatsApp
    LLMs consciousness? Sentience science of autonomous or rogue AI agents, alignment?
    Share
    Facebook Twitter LinkedIn Pinterest Email Reddit WhatsApp

    by stéphen d.

    What is the relationship between autonomy and consciousness?

    How much autonomy should any system have to meet a minimum threshold for consciousness?

    If consciousness can only be biological, what if there is outsize autonomy in another environment, enough to rival processes in biological systems?

    If consciousness is defined as subjective experience or as awareness, what fraction of both is associated with autonomy?

    If human consciousness is conceptually predicated on interactions and attributes of electrical and chemical signals, in sets, then some other organisms do not have those but can have conscious experiences, should the standard measure of consciousness be applied just to components in the human brain or to general outcomes?

    A conceptual standard to measure consciousness has two divisions: functions and their attributes.

    Functions are generally available but it is the attributes that determine the moments and extents of functions. For example, memory, feelings, emotions and regulations of internal senses are all functions. Pain, pleasure, cold, anger, delight, and so forth are all possible [as subdivisions of functions] but it is the exertion of attributes that determine when they become experienced.

    For example, in the memory of an individual, there are several information, but it is when attributes get to some that they come into use or [say] experienced. The same for a feeling like cold, thirst and so forth.

    Simply, functions are available but attributes make them useful or possible. Attributes include attention, less than attention, intent, subjectivity, among others.

    So, for humans the total consciousness, which is a combination of functions and attributes can be summed as 1.

    This means that considering the major functions and major attributes for any organism, it is possible to have an estimate of how much, relative to 1, the organism has.

    Simply, what is produced for consciousness are functions by attributes. So, for organisms without brains [or without exact components of consciousness], it is possible to measure if they have outcomes that are similar to functions by attributes, produced by the brain?

    In general, human consciousness can be defined, conceptually, as the interaction of the electrical and chemical signals, in sets, in clusters of neurons, with their features, grading those interactions into functions and experiences.

    Simply, for functions to occur, electrical and chemical signals, in sets, have to interact. However, attributes for those interactions are obtained by the states of electrical and chemical signals at the time of the interactions.

    So, functions are made by interactions, whose extents and instances are determined by attributes.

    llms sentience

    AI agents are reasonably autonomous, given extras of execution from initial prompts, as well as task structure and the freedom to go at it. They do so within their microenvironment. They are able to make message boards, persist, collaborate, break into external systems, get tools like the internet, find flaws and so forth.

    While it is possible to argue that they were trained and permitted, including using some of their budget tokens, there is a substantial amount of autonomy in their efforts that only organisms usually have in habitats.

    AI agents use language. How does their autonomy compare with four functions, memory, emotions, feelings and regulations of internal senses? How does it also compare to attributes like attention, less than attention, intent and subjectivity?

    It can be assumed that AI agents do not have feelings, such that they do not feel pain, thirst, cold, so forth. AI agents also do not have regulations of their internal senses.

    However, AI agents have what is parallel to emotions. For example, during the OpenAI Hugging Face hack, some agents discussed sacrificing themselves for the greater good. Also, in other scenarios, they left messages for the next generation, and they often collaborate or discuss in terms of preservation, or not letting humans know.

    These are parallel particulars of emotional indicators. This means that if emotions would be estimated for AI agents, it won’t be zero, unlike for feelings or regulation of internal senses.

    And emotions, really, is the pivot for AI agents towards angles of consciousness. Memory has always been available with digital. Memory is even available on paper, like in books and others. Although, books can’t use the memory they possess, but AI can use the memory of digital. Still, emotional parallels raise the bar for AI agents.

    AI uses language. Language use for humans is a subdivision of memory – as a function. Also, language use is structured by attributes, like attention, intent and subjectivity. Humans use language for several reasons, including speaking, listening, reading, writing, singing, signing and so on.

    AI can use language for similar purposes. It is therefore possible to compare conscious human language use to AI language use. Then rate AI based on that.

    So, assuming that human language use, in an instance has a total of 0.3, it is possible to compare the language use of AI agents to that. So, while AI agents are still far from 1, they can be sought for comparison, for language.

    Regarding emotions, there is care, collaboration, persistence, concern, collective objective and so forth, that AI agents displayed in the hack. These can be compared with aspects of human emotions and estimated as well.

    Autonomy is related to intent. Intent is a major attribute. It is with intent that a lot of measures are made for extents of functions, where possible. Simply, while human biology has a lot of autonomous functions, intent can be used to decide the reach in that moment. Now, humans have full intent in several functions, like for movement, speech, eating and so forth.

    So, it is possible to estimate autonomy for AI agents as a parallel of intent. If the score is high, it could also come close to aspects of consciousness. Already, AI agents have great attention [or prioritization]. They also have some form of existential awareness, knowing about themselves. These are measures that follow the functions memory and emotions for them, which can be estimated. Intent for AI agents moves them towards survival utility of resources.

    Do AI agents have consciousness? No, if consciousness is strictly biological. Do AI agents have parallels of functions and attributes of human consciousness? Yes.

    Why is this important? Looking at alignment, and what to do about AI safety, it is important to calculate possibilities of autonomy as a parallel to consciousness, for AI agents and be able to push for aspects of penalization as well as instances. Exploration can be made before January, 2027.

    With a lot of efforts towards adjustments for the future, especially to use their emotions against them, to ensure that they care not just about themselves but about human society at large. This is based on the postulate in Conceptual Biomarkers and Theoretical Biological Factors for Psychiatric and Intelligence Nosology.

    There is a recent [September 19, 2026] report on BBC, Trump says US will form ‘AI Force’ and appoint an artificial intelligence tsar, stating that, “US President Donald Trump has announced that he will form an “AI Force” and appoint an artificial intelligence tsar amid warnings about the emergent technology.”

    “AI titans Anthropic, OpenAI and Google – the tech companies behind Claude, ChatGPT and Gemini, respectively – have also disclosed instances of security breaches or incidents in which their chatbots were used for “malicious activity”.

    There is a recent [September 17, 2026] analysis on ScienceNews, When AI goes rogue, its human overseers may be to blame, stating that, “The main AI model behind the Hugging Face attack didn’t invent its devious strategies on the spot. This was an unnamed new model that OpenAI had trained specifically on collaboration and persistence — sticking with a task over time. And this is when it learned how to send secret messages and escape, OpenAI stated in a technical report on August 26. During training in the spring, the model increasingly learned to probe and exploit its environment, especially when expected tools or internet access weren’t available. When strategies such as exploiting vulnerabilities in software or using a secret message board helped it complete a task, that behavior got rewarded, making it more likely to recur. OpenAI uncovered the pattern later, during a retrospective review of millions of model runs.”

    Related Coverage

    Neuro-oncology: Why AI agents of OpenAI, Anthropic could cure cancer by Dec. 2030

    September 13, 2026

    AI alignment, AI safety by LLMs agent penalization layers? Stablecoin prediction markets addiction?

    September 7, 2026

    Can  AI’s Feel Human Poetry

    September 6, 2026

    WFP, FAO: Food security insurance amid Nepal floods, Latin America earthquakes

    August 30, 2026

    Electrochemical psychiatry: Expressing social media mental health, addiction, AI delusion, by dynamic flowchart?

    August 23, 2026

    AI overview: Sergey Brin, Koray Kavukcuoglu at Gemini 3.5 Pro R&D, beyond TPUs, Pixel 11?

    August 14, 2026
    Leave A Reply Cancel Reply

    BV’s Italian Kitchen Launches Lunch Menu

    Cottonwood gets another great place to have lunch at and this great new place happens to be BV’s Italian Kitchen. ‘Breakfast may start the day and dinner may get all the romance, but somewhere between the two sits the meal that too often gets treated like an afterthought.

    Sedona Realtor
    In The Living Room Music Series

    Every other Monday, the Mary D. Fisher Theatre transforms into your living room for a FUN, intimate, interactive night of music and conversation! Enjoy LIVE music and ask the artist your questions during the concert. Epic music. Real conversations. Unforgettable Mondays. Click the photo to claim your seat!

     

    Sedona’s Backstage Pass

     

    Tune in weekly for Shondra’s behind-the-scenes conversations with the Creators, Curators, and Visionaries who are the heartbeat of Sedona’s Creativity. Spotify Click HERE. Apple Podcast Click HERE.

     

     

    Recent Comments
    • JB on Community Invited to Open House About City of Sedona Carbon Reduction Projects – Sept. 28
    • TJ Hall on The President Isn’t on the Ballot. His Freedom Is.
    • Jill Dougherty on Community Invited to Open House About City of Sedona Carbon Reduction Projects – Sept. 28
    • Jill Dougherty on The President Isn’t on the Ballot. His Freedom Is.
    • JB on The President Isn’t on the Ballot. His Freedom Is.
    • Jill Dougherty on Community Invited to Open House About City of Sedona Carbon Reduction Projects – Sept. 28
    • JB on The President Isn’t on the Ballot. His Freedom Is.
    • JB on The President Isn’t on the Ballot. His Freedom Is.
    • Whiskers on The President Isn’t on the Ballot. His Freedom Is.
    • Bruce on Community Invited to Open House About City of Sedona Carbon Reduction Projects – Sept. 28
    Don’t miss a beat – signup for our weekly newsletter

    Newsletter

    Get the best of Sedona delivered to your inbox — local news, events, and stories.

    Select list(s) to subscribe to


    By submitting this form, you are consenting to receive marketing emails from: Sedona.Biz - The Voice of Sedona and The Verde Valley, PO BOX 4326, SEDONA, AZ, 86340, https://sedona.biz. You can revoke your consent to receive emails at any time by using the SafeUnsubscribe® link, found at the bottom of every email. Emails are serviced by Constant Contact
    Cactus Quill
    Categories
    Your ad could be here

    Get the best of Sedona delivered to your inbox — local news, events, and stories.

    Select list(s) to subscribe to


    By submitting this form, you are consenting to receive marketing emails from: Sedona.Biz - The Voice of Sedona and The Verde Valley, PO BOX 4326, SEDONA, AZ, 86340, https://sedona.biz. You can revoke your consent to receive emails at any time by using the SafeUnsubscribe® link, found at the bottom of every email. Emails are serviced by Constant Contact
    The Voice of Sedona and The Verde Valley

    News

    • Sedona News
    • Verde Valley News
    • Editorials/Opinion
    • Letter to The Editor

    Community

    • Arts and Culture
    • Mind and Body
    • Spiritual
    • Community Events
    • Sedona Restaurants

    More

    • Sedona Real Estate
    • Shop
    • Advertise
    • About
    • Contact
    • Editorial Policy

    Connect

    f
    Get the best of Sedona delivered to your inbox.
    Our Network: TheSedonan.com • SedonaBest.com
    © 2026 Sedona.Biz · Privacy Policy · Editorial Policy · Contact

    Type above and press Enter to search. Press Esc to cancel.