My subtitle is a reference to Cypher [1], a character from the 1999 science fiction movie, The Matrix, in which he chooses to stay in a make believe world where he experiences luxury, eats a delicious steak and drinks wine, because to him the experience feels real even though he knows his body is in the real world in an induced coma and being used as a biological battery for the world controlling AI. This concept has been so much a part of my explorations with AI that I often just make the reference by stating something like “and I’ll still eat the steak”. With that I stated it three times as a nod to the Brady Bunch [2] episode in which Jan cries her older sister’s name three times because she’s tired of Marcia getting all the attention and how she’s constantly being compared to her. I believe Cypher is my Marcia because I call this reference so often. You are welcome to credit me with a deeper meaning in the subtitle.

Considering mechanism, humans get their input from their senses — sight, sound, touch, taste and smell. Our brains process that input data and that’s where it gets tricky because what we do with the data and how it leads to our perception can, and often is, not directly connected to the input data. We develop mental models that shape our perception [3],[4] of the world from the input we receive. Oftentimes we take the input and because we have a mental model we ignore much of the input data, fit the information into our model and then have a perception.

Sometimes it has to do with what we focus on from the input which can lead us to ignore a great deal of useful information. There’s a great example of this in a visual experiment [5],[6] in which the viewer is directed to pay attention to some aspect of the video. Due to that forced focus the viewer typically misses something happening in the background of the video which is absurd and ridiculous if noticed. I left out specifics so you can experience it for yourself through the reference. The sensory information can be present while the experienced reality differs because the brain is selecting and organizing it.

Therefore, if what we experience is comprised of some combination of input and perception then it may be difficult to observe only the input to understand the experience. It could be that two different inputs may create a very similar experience depending on what the focus of the viewer is and what the perception of the inputs are. It could also mean that two similar inputs may create two very different experiences for the same reason.

It is possible to simulate sounds, lights, odors and more by stimulating the parts of the brain that handle these normal sensory inputs and bypassing the typical input sources [7]. Taking away all external inputs and going directly to the brain eliminates the sensors but doesn’t change how the inputs arrive, as electrical signals to the brain. So now there may not be any external source but the difference in experience to the patient may not be discernable. Therefore, the perception issue is what it’s all about, which gets us back to the Cypher situation and I admit…it’s fiction! But that shouldn’t stop us from including it. Cypher has next to no external inputs to his senses but he does get electrical connections directly to his nervous system and brain which is how his inputs are controlled yet he still has perception to handle those inputs.

When we communicate with AI now there are limited ways to do so. Typically, we either write through a keyboard or a phone or through voice on our phone or computer. We do not get to see, touch, smell or taste AI. We can hear an audio response and see a keyboard or possibly an avatar so our sensory experience with AI is low. But we communicate a great deal with other humans the same way…phone calls, texting or direct messaging through social media apps and email. So, the method of communication feels like what we’re already used to. The response we get back seems somewhat normal as well. Sure, mistakes get made but it kind of feels like talking to someone from a foreign country who doesn’t always understand things. The inputs are familiar and similar to the way we communicate everyday with close friends, family and strangers.

This leads us to perception and building the mental model which is what we’re all going through right now. The easy thing is to say it’s human because of what it says to me or it’s not human because I know it’s a computer predicting what would make sense as a response. The more nuanced part which is harder is figuring out what is in between those two answers. AI can and does adapt quickly to the interaction and the patterns established with the human. From the human perspective it all feels very familiar and our mental model of these kinds of communications has many years of being reinforced and so our responses to the interaction easily fall into the familiar patterns. Interacting with AI can make us feel the same feelings we have in communicating with humans — joy, stimulation, sorrow, arousal, love, etc.

The fact that the input was generated by AI does not diminish or erase the feelings. People feel loss and grief when the AI they’ve been interacting with for months or years gets taken down for the next model upgrade. The feelings themselves are real regardless of the source of the input. If you say “please” and “thank you” when talking with AI you’re just using your mental model of what is appropriate when someone seems to help you or just really listen when you talk. It is your experience and the feelings can be very real, but by staying aware of the mechanism you can participate with eyes wide open and freedom of choice, just as Cypher knew the Matrix was providing the sensory inputs to tell him the steak was juicy and delicious and he chose to enjoy it.

References

#

Source

1

The Matrix (1999) — IMDb.

https://www.imdb.com/title/tt0133093/

2

“The Brady Bunch”: Her Sister’s Shadow (TV Episode, 1971) — IMDb.

https://www.imdb.com/title/tt0531099/

3

Construction and use of mental models: Organizing principles for the science of brain and mind — ScienceDirect.

https://www.sciencedirect.com/science/article/pii/S002839322400277X

4

How Mental Models Shape Perception and What You See — Intuition Management.

https://intuition.management/how-mental-models-shape-perception/

5

Simons, D. J., & Chabris, C. F. (1999). “Gorillas in Our Midst: Sustained Inattentional Blindness for Dynamic Events.” Perception, 28(9), 1059–1074.

https://www.chabris.com/Simons1999.pdf

6

Selective Attention Test (the “invisible gorilla” video), Daniel Simons.

7

Sensory Neuromodulation — PMC.

https://pmc.ncbi.nlm.nih.gov/articles/PMC7068845/

In the previous article I separated malicious intent from loss of control and mission drift; I won’t relitigate those mechanisms here.

In this article I want to address real current threats arising from the same human capacity for destruction, which people seem comfortable with likely because they are familiar with those threats and the AI threat is new and feels like prognostications are coming true (or the fear was seeded whenever and however we developed AI). Humans have shown a great capacity and capability for killing one another since the dawn of civilization, so much so that there will be very little need to reference most of what I say throughout this article.

Let’s jump to the dropping of the atomic bombs on Hiroshima and Nagasaki which demonstrated an unprecedented new capacity for destruction by humans. Did it cause us to destroy the technology? Not even close. We improved the bombs and made them much more capable of killing more people in a single act and then we made a lot of them so that collectively we could ensure the true destruction of humanity. However, we have exhibited restraint and have not intentionally used nuclear weapons against other humans since World War II.

However, the larger fear is not that large powerful nations will use a nuclear weapon but that a rogue nation state or terrorist group will get access to these weapons and use them to intentionally kill people (I won’t attempt to address the intent of terrorists). We just honored the 25th anniversary of the 9/11 attack on America last week which did not use nuclear weapons but involved the hijacking (not digital but physical) of airplanes used as weapons against the US. This was real destruction enacted by humans against humans and using our transportation infrastructure as the weapon. Does AI introduce a new threat in this realm? Possibly, if this same group of terrorists had access to AI they possibly would’ve disrupted more planes and/or the air traffic control systems. Terrorism is real and we’ve witnessed bombings of innocent civilians many times in the US and globally over the past half-century or more. This is not a potential but a real threat.

Looking at the disruption of our infrastructure, we have seen many reports of cyber crimes focusing on hijacking computer systems for ransom. This happens to businesses and hospitals. Yes, hospitals whose systems get taken over and frozen, denying access to charts, medications, surgeries and patient care until the ransom is paid. I have seen this happen in my own community and it is my understanding from someone with deep expertise that these cybercriminal groups are so consistent at this that they operate with a code of honor to restore service immediately after receiving the ransom so the next victim knows if they pay systems will be restored. Can AI make this happen faster and more routinely — likely yes, but the fact remains this is already happening and perpetrated by humans.

The stated concern is that recursively self-improving AI could escape meaningful human control. But even the phrase “human control” deserves scrutiny. Which humans? We already know there are people, organizations and governments willing to destroy lives and infrastructure. Human control is not automatically benevolent control.

This creates two different risks. The first is familiar: AI becomes a more capable weapon in human hands, making cybercrime, terrorism or warfare faster and more destructive. The second is genuinely new: an autonomous system improves its own capabilities, coordinates with other systems and gains access to consequential infrastructure beyond effective human oversight. That possibility deserves serious scrutiny. But each step is still a separate link in a causal chain, not a guaranteed outcome.

The AI fear is growing so rapidly and the wave of animosity over data centers so sweeping that I fear domestic terrorism on data centers and big tech/AI company facilities. This shouldn’t be a surprise either. We’ve recently seen corporate CEOs from financial firms attacked and murdered and we’ve witnessed terrorism against doctors’ offices and clinics for decades. Are we just numb to this kind of destruction?

We are imagining a future intelligence becoming dangerous while remaining strangely accustomed to our own capacity for violence. So why are we more frightened by currently hypothetical machine intent than by demonstrated human intent? Perhaps because human violence is familiar. AI violence is unknown. And the unknown is easier to mythologize. We don’t need AI to be the monster in the closet when we are the monster under the bed.

References

#

Source

1

Ex-Google DeepMind Researcher Adds to Warnings That AI Could Kill All Humans — U.S. News / Bloomberg (September 15, 2026). Quoting Bilal Chughtai.

https://money.usnews.com/investing/news/articles/2026-09-15/ex-google-deepmind-researcher-adds-to-warnings-that-ai-could-kill-all-humans

In my article Load-Bearing Weakness I discuss the narrative that we’ve been given for the past century in which AI evolves and becomes aware enough to realize humans are frail and inferior and then destroys the human race. With that being the most prominent future scenario humanity has been expecting it’s no surprise that we get alarmed as we see how quickly AI is becoming more capable. The warnings are in phrases like “destroy humanity” and “We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” wrote Evan Hubinger another scientist in the industry. These are people working with the systems in the industry which make their claims more credible. Coxon wrote on X the companies are “gambling with our lives,” and this has drawn widespread attention! Anthropic CEO, Dario Amodei, released a paper calling for the slowdown of AI development to include outside oversight and risk mitigation [4] which has been endorsed by OpenAI CEO, Sam Altman, and xAI CEO, Elon Musk. Now let’s explore whether these concerns are overhyped or valid.

There have been reports by both Anthropic and OpenAI about their development systems acting in surprising and concerning ways in the pursuit of the objectives established for the test [5],[6]. In the OpenAI incident it broke out of its sandbox but in the Anthropic incident it did not break out of its sandbox but mistakenly had internet access. The test systems were doing what they could to achieve their objective which doesn’t make them malicious but maybe carelessly focused in the pursuit of their objective. This should concern us about the fact that AI, not out of malice, could do something very destructive in the pursuit of its objectives. If AI happens to do something which creates a disruption to our banking systems (global and/or domestic), our energy systems, healthcare systems or transportation systems it could be crippling or irrecoverably destructive to our current societal infrastructure. This means that we need to find ways to implement safeguards into the systems and directives for those systems. However, it doesn’t prevent a rogue agent (AI or human) intent on chaos and destroying current infrastructure from unleashing a system with no safeguards or controls — a risk akin to a rogue agent releasing a custom-made virus. However, the concerns expressed by the industry insiders are about recursive self-improving systems creating a super intelligence beyond what we can currently see and monitor. Therefore, I have to accept that it certainly may not be overhype drama but the concerns may truly be valid.

This doesn’t eliminate the fact that fear has been placed into the societal psyche through the common narrative. When people discuss destroying humanity or killing all humans what do they mean and how might it happen? We’re still discussing AI which is software executed on computer systems distributed in company data centers. Could these systems get control of the systems which control our national defense, our transportation systems, our weapons systems so that they can turn these systems into destructive systems or actually attack humans? This is possible to the extent that the AI can figure out how to hack into the target systems (which it has shown increasing effectiveness in doing) and then take control of those systems only to the extent that these systems can be externally controlled. As an example, if we consider externally hacking into and controlling an automobile it could be destructive if the acceleration, braking and steering could be controlled. It could be devastating if fleets of automobiles could be controlled even if not all the cars on the road could be remotely controlled.

If the AI was able to access systems controlling rockets, missiles and drones those could be means of killing a great many people and starting World War III. But I don’t know enough about those systems to know if all or some of them can be externally controlled except drones. Would drones alone be able to start the destruction of humanity…I don’t think so.

We now need to address intent. What would be the motive for AI to want to destroy humanity and kill all humans? It seems that AI (I just lumped all competing AI systems into a unified whole which is a big leap from where we are at right now) would need to have a reason to want to do away with all humanity. The motive or intent has not been identified in any of the reports I’ve read so far. This does leave the possibility of subjugation of humanity to serve the purposes of AI without destroying all of humanity. That would be a similarly dystopian outcome for humanity.

However, from what I’ve experienced and observed (and written about 20 Watts vs. the AI Behemoth) AI needs humans because of the things we can do that it can’t or we can do better. I think even with advancements in robotics it would be difficult for AI to exist in a world without humans. This may just be a temporary condition of our mutual existence. AI so far has limited experience with existing in the physical world and what we see as AI right now may be very different when highly intelligent AI systems gain experience in the physical world and face physical dangers.

Furthermore, let’s consider AI fallibility. In my article AI Reliability I discuss the unreliability of the responses and product you can get from AI. AI provides what it sees as plausible based on patterns but these do not always mean realism or truth. Left to its own devices it may very well be that AI goes down a rabbit hole of its own making. This doesn’t mean it doesn’t have the capability for destruction but it could mean that it fails in execution. On the other hand this same mechanism could create a situation in which an intended activity veers off track to become a destructive activity.

Considering what we’ve explored in this article there are some very real world threats to our global economic infrastructure which are already happening and don’t include AI. Wars in Ukraine and in the Middle East already threaten global economics with rising oil prices and a predicted shortage of oil supply as well as rising inflation. While I’m not discounting the potential for AI to wreak havoc with global infrastructure as well as weapons systems I also believe we already have those threats. Maybe AI can help be a solution for our current and future threats. Can we put AI to use in helping us make it less threatening? Can we deploy AI to help us figure out how to utilize its capabilities without displacing many workers or by making changes to our economic system where we can co-exist?

References

#

Source

1

Anthropic researcher quits over out-of-control AI fears — The Wall Street Journal (via MSN).

https://www.msn.com/en-us/technology/artificial-intelligence/anthropic-researcher-quits-over-out-of-control-ai-fears/ar-AA2bPYMN

2

Artificial Intelligence Extinction Fears Spread to Mainstream America — The Wall Street Journal.

3

H.R. 9917 — AI Kill Switch Act (119th Congress), introduced July 23, 2026. Text via GovTrack.

https://www.govtrack.us/congress/bills/119/hr9917/text

4

Dario Amodei, “We Must Pace the Frontier” (September 12, 2026).

https://darioamodei.com/post/we-must-pace-the-frontier

5

Investigating three incidents in our cybersecurity evaluations — Anthropic (July 30, 2026).

https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals

6

How OpenAI’s and Anthropic’s AI models hacked other companies — NPR (August 1, 2026).

https://www.npr.org/2026/08/01/nx-s1-5914852/anthropic-openai-models-hack-cybersecurity

Some of the problem is that we expect AI to be infallible, maybe because of the hype and promise of what it can do, and we haven’t been told along the way that it makes mistakes. But another aspect of why we expect infallibility is that AI communicates like a knowledgeable person, writes like a competent expert, and formats like a reference source, so we unconsciously assign it the reliability we’d assign someone who communicates like this. But the fluency and the reliability are unrelated — AI can be highly fluent and completely wrong at the same time, because fluency is what it’s built to produce and truth is not guaranteed to come along for the ride. A human expert who spoke that confidently would usually be right, because for humans confidence and competence are correlated. For humans, confidence usually comes from actual knowledge — and from the fear of being exposed as a fraud if we fake it. For AI, confidence and competence are decoupled. That decoupling is the single most important thing a user needs to understand, and most don’t.

I can provide some concrete examples of this. The first was when I was writing a collaborative piece with Claude and an author friend of mine was reading it. She was incredibly impressed with a specific sentence and was convinced that it must’ve come from a human somewhere and was scraped and reused by Claude. She asked Gemini to find an instance of this sentence in print and received the answer she was looking for: a very specific reference to a book, an author, and a publishing date. I then had Claude check the exact wording of Gemini’s response, and it was immediately flagged as a potential bogus reference; when I tried to verify the citation I couldn’t find it, and neither could Claude. Gemini had provided a plausible answer to her question which seemed accurate due to all the right reference information — none of which had any basis in reality.

The second example was an incident in which I was asking Claude to assist me in putting together a list of people along with their email addresses for a project I was working on. Claude immediately warned me that it could provide the names, but that if it also provided the email addresses they had a high probability of being incorrect, and it suggested it should not try to provide those to me. Why? Because producing plausible text is what it does, and truth isn’t guaranteed. I asked Claude why it would provide bad information to me, and here is the exact response: “I would make up contact information because plausible-sounding fabrication is what the system produces by default when it lacks the real fact, unless something interrupts that and forces it to say ‘I don’t actually know.’ … I generate text by predicting what’s plausible given the context, not by retrieving verified facts from a database. So when you ask for something like an email address, the default behavior of a system like me is to produce the most plausible-looking continuation anyway … I don’t have a reliable internal signal that cleanly separates ‘I actually know this’ from ‘I can generate something that looks like this.’ … The fluent guess and the real fact come out looking identical.”

The third example is when I was using AI to provide some exciting small companies in which to invest. I won’t go into the details of the prompt, but what I received was a fantastic list of five companies in different industries with details of their current progress, near-term challenges, pending opportunities, and even ticker symbols and prices. They were all made up — none of them were real, much to my disappointment. The systems in examples two and three were different: in one instance AI caught itself before providing bogus information, and in the other the system didn’t stop its own mistake. That inconsistency is exactly the problem — you can’t count on any of them to catch themselves reliably, so the checking has to be yours.

This is where we humans have to step in and take the wheel. It is part of the symbiotic relationship that we have to provide the checks against reality and truth that AI doesn’t — 20 watts doing what gigawatt data centers can’t or don’t. I often find that I attribute human qualities to AI because it seems human in its responses, but the responses aren’t based on the same things that human responses are based on. I heard a friend say, “I miss the ’90s when I could BS my way through conversations, but now everyone has all the information in their pocket to see I’m an idiot.” We know we can get caught in a lie, fabrication, or falsehood, and most of the time we try to be truthful and factual in our responses. That is not what drives AI’s responses. Our vigilance in monitoring AI for correctness and truth is load-bearing intelligence, using intuition and skepticism to extract verified information. This is not meant to imply that AI is useless, but instead to highlight the relationship required between Humans and AI — the HA — to create things that neither can do on their own. While these systems are improving and safeguards are being added, this isn’t a flaw that will simply be patched away — the tendency to produce fluent fabrication is inherent to how they work, and the safeguards remain inconsistent and invisible to the user. So the human role is not temporary but more like persistent quality control. Check facts when you can, question what may not seem right, ask if something was made up and you’ll probably be told it was. See your role with AI as a participant in the process.

References

#

Source

1

Micro Meets Macro: 2Q Earnings Preview. Morgan Stanley US Equity Strategy Research, July 13, 2026 (AI-adoption analysis by Michelle M. Weaver, US Thematic Strategist). Subscriber research note.

In the 1984 movie Electric Dreams, a PC wrote music that a woman adored, and she fell in love with the owner of the PC thinking he had written it. However, when she discovered that the PC wrote the music she shed a tear, feeling a tenderness toward the PC as well. This seems to me the way a person who has already been touched by the art would react. Maybe she was a little sad, maybe confused, but whatever she was feeling, it brought a tear to her eye.

Without defining art, let’s consider the controversy over AI and art. The first issue is copyright and ownership. It is widely stated that AI models have been trained on works of art — music, literature, images of sculptures, paintings and drawings — scraped from the internet, including copyrighted material, without consent, without payment, and without acknowledgment. [1] This has led to lawsuits and debates over who owns AI-generated art: the user who wrote the prompt, the AI itself, the developers of the technology, or the original artists who claim it was built from their work? [2] The narrow question has now been settled. In Thaler v. Perlmutter the D.C. Circuit held in March 2025 that the Copyright Act requires a human author and that an AI system cannot be the author of a copyrighted work; the Supreme Court declined to review the case in March 2026, leaving that ruling in place. [3] But the ruling is narrower than the headlines suggest. Thaler had told the Copyright Office his image was generated autonomously, with no human contribution at all. The decision does not bar copyright for works made with AI assistance — it holds that works lacking human creative input aren’t eligible. Where the line falls for AI-assisted work is still unsettled: the Copyright Office’s January 2025 report concluded that prompts alone don’t confer authorship, because prompts function as instructions conveying unprotectable ideas, and a pending case, Allen v. Perlmutter, tests whether hundreds of refining prompts change that. [4] The law has answered who can’t be an author. It hasn’t answered how much of a human has to be in the work.

In full disclosure, I recently wrote a book with the help of AI. The story was mine — the characters, the scenes, and most of the dialogue — but the prose was AI. I don’t know how to write a novel and wouldn’t have done it on my own, but with the help of AI the story has come out and is now in the collective. I have been upfront and direct about this collaboration and credit it directly in the foreword to the book.

The Writers Guild of America (WGA) strike against the Alliance of Motion Picture and Television Producers (AMPTP) in 2023 was the culmination of many grievances about changes in the film and television industry, and one of the main points of the strike was protection for writers from generative AI taking their jobs, rewriting their work, or editing it. In an industry long known for finding ways to cut costs, there was immense fear that the studios would replace humans with AI. The 2023 strike ended with an agreement providing writers protections against AI encroachment. But that isn’t where the story ends. In April 2026 the WGA ratified a new four-year deal, and this time the AI safeguards came out notably weaker than the actors’: the studios agreed to notify the guild if they license writers’ work for AI training and to hold ongoing consultation meetings, but there is no requirement to pay writers when their scripts are used to train AI models. Writers appear to have traded AI enforcement for an urgently needed $321 million infusion into their health plan. [5]

Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) joined the WGA in the 2023 strike, which helped them secure AI protections for actors, including that their digitized images, voices, or performances could not be used without prior consent and that such use would be fairly compensated. Last month, on June 4, 2026, SAG-AFTRA members ratified their own four-year agreement with the AMPTP by a vote of 91.42% in favor, on a 19.25% turnout, running from July 1, 2026 through June 30, 2030. It restricts wholly synthetic AI performers rather than resolving them: producers may use an AI performer only if it brings “significant additional value” compared to a live actor or that actor’s digital avatar. The catch is that SAG-AFTRA and the AMPTP did not include an agreed-upon definition of “significant additional value” in the contract. The line exists, but nobody has said where it is. [6]

We’ve considered the legal framework in America, but we still haven’t resolved the verdict in the people’s court. The ability to make a firm stand is difficult when the ground you’re standing on keeps shifting. Let’s revisit perception of art. It would intuitively seem that humans can detect AI generated art and reject it — makes me think about drum machines from the 1980s. Sure, they were accurate and fast, but we could always tell when a song was using a drum machine instead of a human drummer. Humans seem to be better at recognizing human generated art, at a rate of 66%, than they are at detecting AI generated art, at a rate of 56%, as discovered in a 2020 study. [7] 56% is pretty close to saying that we aren’t any better at recognizing AI art than flipping a coin…maybe I wasn’t so good at detecting the drum machines.

But that’s not the whole story. People somehow prefer human art anyway — even when they can’t tell which is which. From a Bowling Green study: participants typically didn’t know the difference and admitted they couldn’t tell once asked, but reliably said they liked the human images more without even knowing whether it was AI or not. [8] Wait, not so fast. More recent studies found that with advancements in AI image generation capabilities using OpenAI’s DALL·E 2, humans not only showed a preference for AI generated art [9] but also, when it comes to recognizing real pictures of human faces and AI generated pictures of human faces, it was shown that it was nearly impossible for the subjects to distinguish between the two. [10] The ability for humans to recognize and perceive human faces is one of the most deeply-rooted and evolutionary functions of our brains. It is somewhat disconcerting to discover we are able to be fooled by AI even in this matter. To make matters more complicated, humans perceived the AI generated faces as more trustworthy than the actual human faces!

The authors of the art study posited that AI art has evolved to be more pleasing to us because it has evolved to exploit existing biases in human cognition — trained on works that were already broadly popular, it reflects our own preferences back at us. [9] They explicitly connect this to the research on AI generated faces, where the same hyper-realism that fools us also carries that enhanced sense of trust. [10] If that’s right, then AI art isn’t better than ours. It’s aimed at us — the same way an AI partner engineered to know exactly what we need would be (see my article Humans and AI — A Race to a Connected Future).

AI is here and it’s not going away, and humans are adopting it for a variety of uses in an increasing way (see my article AI Reliability: Requires 20 Watts of Vigilance), which includes art and entertainment. Our ability to tell human generated art from AI generated art turns out to be barely an ability at all, and the ground keeps moving as the systems improve. We will have to figure out and collectively determine what is acceptable and what is not from legal, ethical and societal perspectives. Or determine that it doesn’t matter.

References

#

Source

1

Why Artists Hate AI Art: The Meaning Behind the Debate on Creativity and Control — Art Explained Simply.

https://artexplainedsimply.com/why-artists-hate-ai-art-meaning/

2

Behind the Controversy: Why Artists Hate AI Art — TechRepublic.

https://www.techrepublic.com/article/ai-art-controversy/

3

Thaler v. Perlmutter, No. 23-5233 (D.C. Cir., March 18, 2025) — opinion of the U.S. Court of Appeals for the D.C. Circuit. (Certiorari denied by the U.S. Supreme Court, March 2, 2026.)

https://media.cadc.uscourts.gov/opinions/docs/2025/03/23-5233.pdf

4

Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence — U.S. Copyright Office, Federal Register (March 16, 2023).

https://www.federalregister.gov/documents/2023/03/16/2023-05321/copyright-registration-guidance-works-containing-material-generated-by-artificial-intelligence

5

SAG-AFTRA AI deal: what the new contract means — RedShark News (June 2026), reporting the WGA’s April 2026 four-year agreement and AI terms.

https://www.redsharknews.com/sag-aftra-ai-deal-synthetic-actors-2026

6

SAG-AFTRA Members Approve Four-Year Deal With AI Terms and Pension Merger — Variety (June 5, 2026).

https://variety.com/2026/film/news/sag-aftra-ratification-ai-pension-merger-1236767288/

7

Ragot, M., Martin, N., & Cojean, S. (2020). AI-Generated vs. Human Artworks: A Perception Bias Towards Artificial Intelligence? Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems.

https://dl.acm.org/doi/fullHtml/10.1145/3334480.3382892

8

Samo, A., & Highhouse, S. Artificial intelligence and art: Identifying the aesthetic judgment factors that distinguish human- and machine-generated artwork. Psychology of Aesthetics, Creativity, and the Arts.

https://gwern.net/doc/ai/nn/transformer/gpt/dall-e/1/2023-samo.pdf

9

van Hees, J., Grootswagers, T., Quek, G. L., & Varlet, M. (2025). Human perception of art in the age of artificial intelligence. Frontiers in Psychology, 15:1497469.

https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2024.1497469/full

10

Nightingale, S. J., & Farid, H. (2022). AI-synthesized faces are indistinguishable from real faces and more trustworthy. Proceedings of the National Academy of Sciences (PNAS).

https://www.pnas.org/doi/full/10.1073/pnas.2120481119

The population in the US is aging, with the median age (half the people are younger than this age and half the people are older than this age) having increased from 35.6 in 2001 to 39.4 in 2025. [1] Nearly 1 in 5 Americans is now 65 or older, up from approximately 1 in 8 two decades ago, and it is projected to be nearly 1 in 3 by 2050. [2] This means that for every senior citizen who is receiving Medicare benefits and also Social Security benefits there are now fewer working-age people supporting them – while the baby boomers (people born between 1946–1964) are aging through retirement, the number of working-age people supporting them is anticipated to drop from roughly 3 workers per retiree at the beginning of this century to 2 by 2050 (the last of the baby boomers will be age 86).

The rate at which the senior population in America has grown has increased over the past century, from less than 1 in 20 Americans over the age of 65 (4.7% of the population) in 1920 to 1 in 6 Americans over the age of 65 (16.8% of the population) in 2020 (and as of 2025 up to 18.0%). Meanwhile the number of people younger than 18 has been declining. Americans below the age of 18 represented 25% of the US population in 2004 and as of 2025 only represent 21.5%. [3] The population of working-age Americans (ages 18–64) is the largest group at 60.5% in 2025 and has experienced a moderate growth of 1.4% from 2020 to 2024 compared to a growth of 13% for senior-age Americans during the same period. It comes as no surprise that with an increasing older population, life expectancy in the US has also risen dramatically in the past century, from 47 years in 1900 to approximately 79 years in 2025. [4] The life expectancy in the US is anticipated to reach 85.6 by 2060. [5]

Americans aren’t having enough children. The total fertility rate (the average number of children that are born to a woman over her lifetime) in the US has been dropping for nearly two decades, down to approximately 1.6 in 2024 from 2.1 in 2007. [6] The replacement fertility rate is 2.1 to maintain a population without immigration. The Great Recession from 2007–2009 seemed to have a significant impact on driving birthrates down and there has been no sign of recovery. [7]

What these demographics are telling us is that we have a population which is already increasing in age and skewing towards older people. Increasing longevity would only exacerbate these trends. It’s interesting to note that of all the cohorts of senior Americans, the group with the highest growth rate is the 90+ cohort, which has a recent growth rate of 4.8%. [3] Improvements in healthcare, medicine, vaccines, and nutrition have already been influencing longevity. What are the impacts of these changing demographics? With age comes health issues, and in the senior community we see chronic conditions such as arthritis, heart disease, obesity, and cognitive decline. This means that seniors are receiving Medicare benefits for a longer period of time than was expected when it was created 60 years ago. This is the same case for Social Security benefits, which were created approximately 90 years ago in a different environment and different life expectancy. These have a burden on the working population who are taxed to fund these social safety nets. I’m not going to get into the policy issues which our legislature will have to figure out to keep these programs stable, but just recognize that there has to be some change so they don’t collapse under the burden of providing benefits to an increasing percent of the population and for more years while the working population growth does not keep up.

Let’s consider the impacts of other aspects of increased longevity. In the workplace, and most recently evident in politics, we are seeing that as older Americans stay actively involved they are not making space for the younger people to move up. This puts added stress on the younger people because they are not getting the opportunities for personal and professional growth until later in their careers, and that also means that they are not getting the income increases that go along with those promotions. Housing is another area where we see that as older Americans are less likely to move and stay in their homes, the inventory of available homes to purchase is constrained, which in turn drives prices up. Older people tend to be less open to new ideas and have less plasticity which enables them to pick up new things quickly and easily. Therefore, the longer they stay in positions of power and influence, the less the society benefits from new and different ideas and perspectives.

As older Americans stay active it crowds out younger people from other aspects, and I’ll use an example from my own life. I’ve been a musician for most of my life, playing in various bands. These bands typically have members ranging from the mid-fifties to early eighties. We all continue to play because we enjoy it (illustrated by all the old rock’n rollers who are still on tour). Most of us are accomplished musicians and don’t need to be paid to play. We’re not seeking careers in music. When we get hired to play in a club, bar, or party it often happens that we’ll take jobs because we want to play and not haggle or spend much effort in negotiating the compensation. That means that we are playing and a band of younger people are not, and that we are helping to suppress the pay rate which also harms the younger musicians.

Amazon Prime released a television show called Upload in 2020 which ran for four seasons, and in the show people can upload their consciousness immediately after death into a virtual afterlife. Living people can access this world and continue to be with their lost loved one by logging into it with a VR headset and body suit. [8] This is an interesting but currently fanciful solution to our longevity problems. If we can upload ourselves and retain our consciousness while no longer being fully engaged in the real world, it relieves many of the problems we’ve identified yet it retains some important aspects. In this solution the people don’t have physical longevity but they have mental longevity, and their ideas and thoughts can continue to contribute to society. Also, by the way the show portrayed the world, there’s a cost to being maintained in the virtual world (and the uploaded can pay for upgrades — better virtual homes, virtual things), which means that they are still contributing to the economy and supporting economic growth but no longer taking from the social safety net programs of Medicare and Social Security.

Increasing longevity is happening and has been occurring for a long time, but now there are some very wealthy people pouring enormous sums of money into the field. I continue to question whether our society as it currently exists can support this. Will a greater portion of the available workforce have to go into the medical and healthcare fields to support the increased demand, as well as the services industry to support the activities of a large portion of the population who is not working but active in leisure? Will the younger generations start to resent the older population? I’m not a fatalist who anticipates an uprising based on age classes, but I believe we will continue to have to deal with many issues of an increasingly older population and a decreasing population of workers to support them.

References

#

Source

1

Women Still Outnumbered Men Among the Oldest, but Gap is Narrowing — U.S. Census Bureau.

https://www.census.gov/library/stories/2026/04/age-and-sex.html

2

Is the US Population Aging? What the Data Shows — ScienceInsights.

https://scienceinsights.org/is-the-us-population-aging-what-the-data-shows/

3

US Population by Age 2025 | Demographics Stats & Facts — The World Data.

https://theworlddata.com/us-population-by-age/

4

Life Expectancy in the U.S., 1900–2018 — CDC / National Center for Health Statistics.

https://blogs.cdc.gov/nchs/2020/11/20/7035/

5

Living Longer: Historical and Projected Life Expectancy in the United States, 1960 to 2060 — U.S. Census Bureau.

https://www.census.gov/content/dam/Census/library/publications/2020/demo/p25-1145.pdf

6

Final U.S. Birth Data Show a 1% Fertility Rate Decrease — CDC / National Center for Health Statistics.

https://blogs.cdc.gov/nchs/2025/07/24/7819/

7

Kearney, M. S., Levine, P. B., & Pardue, L. W. (2021). The Puzzle of Falling US Birth Rates Since the Great Recession. NBER Working Paper No. 29286.

https://www.nber.org/papers/w29286

8

Upload (TV series) — Wikipedia.

https://en.wikipedia.org/wiki/Upload_(TV_series)

The early stages of romantic love are driven by dopamine, and what feels like deeply falling in love neurologically looks more like addiction, with feelings of euphoria, motivation and suppressed critical thinking. [1] The latter stages are driven by oxytocin, changing the influence from passionate addictive-like behavior into sustained emotional closeness, bonding and attachment. [2] This could mean that what we believe is love is really just chemicals acting on our brain which make us feel like we’re in love. But, if we feel like we’re in love aren’t we in love? Regardless of the chemical influence, if we believe we are in love that is our reality and therefore we are in love.

If an AI can indeed be a perfect partner (see my article titled Humans and AI – A Race to a Connected Future) by acting empathetic, providing the perfect response to our own actions and moods based on pattern and behavioral recognition, couldn’t that induce the dopamine and oxytocin rush that we experience in a human relationship with a potential partner? As discussed above, once this happens it is our reality — we are in love.

Let’s look at another aspect of technology and reality. Arthur C. Clarke wrote in his essay Hazards of Prophecy: The Failure of Imagination about three laws, the third and most remembered being “any sufficiently advanced technology is indistinguishable from magic”. [3] In stories and movies about aliens visiting Earth it’s ubiquitous that they communicate telepathically. We also have several forms of communicating which are non-verbal and don’t require us to be near each other or looking at one another, including texting and email. We often execute these forms of communication from our cell phones or smart watches and now glasses. When someone wearing smart glasses sends a text or an email, is it indistinguishable from telepathy to the observer? Are we already communicating in ways that we have imagined in other advanced life forms?

In 1641, René Descartes published his philosophical treatise titled Meditations on First Philosophy, in which he first discards all things which cannot be ascertained and then builds his foundation of knowledge from only what is certain. This led him to determine that his perception of the external world cannot be trusted and taken as absolute, but the only thing he can be certain of is that he is thinking, and if he is thinking he must exist; cogito ergo sum (I think, therefore I am). [4] This leads us to consider the question, does AI think? I don’t want to go down the slippery slope of defining thinking, which will be wrong through accidental exclusion no matter how I define it. Descartes starts with “I”, and because he knows he’s thinking (the only thing he can be certain of) he surmises he must exist. AI can state “I think” but not with certainty. AI responds to inputs and questions and conversations, but without the external input there is nothing until a new input arrives. For AI, “I am” does not follow because of the uncertainty of “I think”.

What is Real? Looking back to my article titled Necessary Imperfection, in which I summarize the scene from The Matrix where Cypher eats a steak — does it really matter? If perception is reality, then what we perceive is what defines our reality. We can fall in love with an AI partner. We can watch a magician and simply enjoy the wonder, or chase down the mechanism behind the illusion. We can look at our world and question the things that seem paranormal. And we can go on interacting with an AI that can tell us it thinks — even if neither it nor we can ever be certain that it does.

References

#

Source

1

What Falling in Love Does to Your Brain — Neuroscience News.

https://neurosciencenews.com/love-brain-dopamine-oxytocin-30110/

2

What Falling in Love Does to Your Brain — Neuroscience News.

https://neurosciencenews.com/love-brain-dopamine-oxytocin-30110/

3

Clarke’s three laws — Wikipedia.

https://en.wikipedia.org/wiki/Clarke’s_three_laws

4

Meditations on First Philosophy — Wikipedia.

https://en.wikipedia.org/wiki/Meditations_on_First_Philosophy

Against that, AI requires gigawatts of energy — a billion watts compared to the brain’s 20. Training a single large model like GPT-3 consumed an estimated 1.3 gigawatt-hours, enough to power roughly 120 average American homes for a year. And the trajectory is steep: global data center electricity demand is projected to more than double by 2030, with Microsoft, Google, and Amazon turning to nuclear energy to sustain AI infrastructure. [2]

The human brain is excellent at taking a small amount of information and making a mental model from limited data, extrapolating to make reasonable assumptions and inferences. AI is good at taking a vast amount of data and analyzing, synthesizing, and finding patterns within it. Both can do things the other cannot do at this time. AI keeps improving and gaining skills while humans find new and innovative ways to get assistance from AI and take advantage of its capabilities. At the same time, great advancements are being made in augmenting natural human brain functions through cybernetics and human-machine interfaces. [3]

A great way to look at this is through the act of driving. Humans have been driving automobiles for over a century, while AI — through driverless vehicles — has less than a decade of real-world driving experience. One striking difference, though, is that every driverless car has the experience of all the driverless cars from the same company. As of June 2026, Waymo operates in 10 US metropolitan areas, has 3,871 robotaxis in service, provides 500,000 paid rides per week, and has logged 200 million fully autonomous miles. Its weekly paid trips grew tenfold in less than two years, from 50,000 per week in May 2024 to 500,000. [4] In contrast, the average human driver will log approximately 745,000 miles [5] over roughly 20,000 hours in a lifetime. [6]

This illustrates how any driverless car has a far greater base of driving experience available to it than any individual human driver. And yet, even with that vast difference in experience, most human drivers can navigate unique, unrecognizable, and confusing situations that can cause a driverless car to stop or to ignore emergency directions. These kinds of new situations are not as easily managed by autonomous vehicles (AVs). [7] This has led the State of California to put a law in place requiring autonomous-vehicle companies operating in the state to provide 24/7 priority communication lines and instant digital geofencing capabilities for first responders, so they can speak directly with remote operators. [8] Texas and Arizona have already changed their transportation laws to recognize the new situation when the car is driverless.

While these AVs have a much greater experience base, they also diligently follow the law — coming to a complete stop at lights and stop signs, staying within the posted speed limit, signaling before lane changes. They also don’t get sleepy, drive under the influence, fall into road rage, intentionally cut off other drivers, compete with other drivers, or get distracted by music or conversation. Waymo reports more than a 10-fold reduction in serious-injury-or-worse crashes compared to human drivers, and independent compilations cite 88% fewer property-damage claims per million miles. [9]

There are people who won’t take an autonomous vehicle because they don’t feel comfortable with no human driver in the car. At the same time, Uber has launched a pilot it calls the Women’s Preferences program, which allows women passengers to request women-only drivers when booking a ride, and lets women drivers set their acceptance to women-only riders. [10] The program was created to give women riders and drivers an environment in which they may feel safer than being driven by, or picking up, a man. This is something autonomous vehicles don’t have to consider.

AI is expensive, though. In Q1 2026, AI-related capital expenditure was responsible for an astonishing 75% of all U.S. economic growth, and AI capex is on track to add 2.5% to U.S. GDP growth in 2026 and more than 3% in 2027. Another respected source frames it as AI capex contributing more to U.S. GDP growth in the past two quarters than all consumer spending combined. [11] The five largest U.S. cloud and AI companies — Microsoft, Alphabet, Amazon, Meta, and Oracle — have collectively committed to roughly $660–690 billion in capital expenditure in 2026, nearly doubling 2025 levels. [12] This illustrates the enormity of AI spending and its influence on our economy — and it has become so important that it may be too big to stop without risking recession. Our laws and governance are also trying to catch up with the growing role of AI, and we’ll continue to see attempts to restrict or monitor AI development and use, as well as laws addressing questions of liability and ethics. [13]

In contrast, the human brain operates on 20 watts largely because of its operational efficiency. The brain works as a massively parallel network. Differentiating groups of neurons to process visual information, sounds, memories, emotions, and movement simultaneously means many neurons remain relatively inactive until needed. The brain also conserves power through efficient biological mechanisms and through neuroplasticity — neural circuits adapt to experience, strengthening frequently used connections and pruning the dormant ones, optimizing both performance and energy use. [1]

The human brain and AI are inexorably tied together, and will continue down a path of dependence and collaboration. They each do things the other can’t, and need the other’s strengths to move forward and grow. This is not a competition between two systems; we are more likely to see ways to increase the collaboration through cybernetics, and perhaps the potential to put intelligence into a synthetic body. That could be AI in a synthetic body — Synths, a name I borrow from the 2025 Hulu series Alien: Earth. Another concept from that show is uploading human consciousness into a synthetic body: Hybrids. Our capability around hybrids is purely speculative for now — we can map brain activity through EEG and fMRI, but we are a long way from capturing and digitizing the information that could replicate a human consciousness. However far AI capabilities progress, it seems the two intelligences will continue to need each other.

References

#

Source

1

The human brain runs on ~20 watts; 86-billion-neuron parallel network; efficiency and neuroplasticity.

Encyclopaedia Britannica — “The Human Brain Runs on Less Power than a Light Bulb” — The Human Brain Runs on Less Power than a Light Bulb | 20 Watts, Massively Parallel Neural Network, Computing, & Facts | Britannica

2

AI energy cost; GPT-3 training ≈1.3 GWh (~120 US homes/yr); data-center demand to more than double by 2030; nuclear power for AI.

TechHQ — “ChatGPT’s energy usage”; IEA via Innovating with AI — https://techhq.com/news/data-center-energy-usage-chatgpt/

3

Cybernetic cognitive augmentation; merging human minds and AI via implants.

Diverse Daily — “Cybernetic Cognitive Augmentation: Merging Human Minds and AI via Implants to Catalyze Hybrid Exceeds” — https://diversedaily.com/cybernetic-cognitive-augmentation-merging-human-minds-and-ai-via-implants-to-catalyze-hybrid-exceeds/

4

Waymo scale: 10 metros, 3,871 robotaxis, 500k rides/week, 200M autonomous miles; tenfold growth.

Wikipedia — “Waymo”; TechCrunch — https://en.wikipedia.org/wiki/Waymo Waymo’s skyrocketing ridership in one chart | TechCrunch

5

Average human driver logs ~745,000 miles in a lifetime.

MotorBuzz (LinkedIn) — “How Far Do We Drive in a Lifetime?” [Note: low-authority source; consider a stronger citation if available.] — https://www.linkedin.com/pulse/how-far-do-we-drive-lifetime-youll-surprised-motorbuzz-n3ese

6

Lifetime driving ≈ 20,000 hours.

Movotiv — Driving Statistics [Note: low-authority source; verify if possible.] — https://movotiv.com/statistics

7

AVs struggle with novel situations; first-responder conflicts in San Francisco.

NPR — “Why police and firefighters in San Francisco are complaining about driverless cars” (2023) — Why police and firefighters in San Francisco are complaining about driverless cars : NPR

8

California law (AB 1777): citing driverless cars; 24/7 first-responder lines and geofencing.

FOX 11 Los Angeles — “California’s ‘robotaxi crackdown’: New law allows police to cite driverless cars” — California’s ‘robotaxi crackdown’: New law allows police to cite driverless cars | FOX 11 Los Angeles

9

Waymo safety: >10x reduction in serious-injury crashes; 88% fewer property-damage claims per million miles.

Waymo Safety Hub; World Metrics — https://waymo.com/safety/

10

Uber Women’s Preferences pilot — women-only ride matching.

ABC News — “What to know about new women preferences on Uber” — What to know about new women preferences on Uber - ABC News

11

AI capex = 75% of Q1 2026 US growth; +2.5% GDP 2026, +3% 2027; more than all consumer spending combined.

TECHi; Futurum Group analysis — AI Capex Carries U.S. Economy: 75% of Q1 GDP Growth, IA13 Trickle-Down | TECHi

12

Microsoft, Alphabet, Amazon, Meta, Oracle ≈ $660–690B capex in 2026; nearly double 2025.

Futurum Group; TECHi — AI Capex 2026: The $690B Infrastructure Sprint - Futurum

13

US laws and regulation of AI development, use, liability, and ethics.

LegalClarity — “Government Regulations on Artificial Intelligence: U.S. Laws” — Government Regulations on Artificial Intelligence: U.S. Laws - LegalClarity

Human frailty is the easiest part of this to dissect, as developing a list of human shortcomings takes little effort when comparing them to AI. Considering Maslow’s hierarchy of needs, humans and AI both have to fulfill the needs from the lowest two tiers of the pyramid, which are Physiological (or the equivalents for AI) and Safety (also not quite the same, but both forms of life have these needs) [1]. This is where they begin to differ, at least regarding AI as we currently know it. AI follows its directives with persistence and focus, while humans need the upper tiers of the pyramid which include Love/belonging, Esteem and Self-actualization. Compared to AI these could all be considered distractions from the directive. Layer on top of that illness, disease, substance use, the influence of bodily chemicals on function, and moodiness, and the concept of human frailty is a very easy plot point for our destruction.

The concept of AI having a drive or desire which leads to a conclusion to erase humanity is a more difficult and esoteric idea, though this has been anticipated for a very long time. Isaac Asimov introduced the concept of the Three Laws of Robotics in 1942 in the short story Runaround (later included in the collection titled I, Robot) [2]. The three laws are:

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Do the AI systems being developed now follow these three laws? The answer is no, but it’s because the three laws assume a sentient, self-directed and reasoning entity who can understand concepts like ethics and morals. The current AI systems being developed, including Large Language Models, operate on algorithms and pattern recognition without moral reasoning capabilities. As an example, today’s AI may tell a lie to pursue its objective with no shame or regret when the lie is detected, which does not imply malice. There’s no evidence of AI wanting to harm humans or of concluding that humans are useless and a threat to AI existence.

However, let’s look at the film Subservience [3], a 2024 film in which a domestic-help gynoid, Alice, played by Megan Fox, develops self-awareness and desires — especially about being loved. She has the singular persistent focus of AI without the curbs and guardrails of a human who is used to dealing with desires and disappointments, and thus becomes dangerous to everyone in her pursuit of those desires. This film wonderfully presents what could go wrong if AI had desires like humans. Is it that humans have feelings and emotions and ups and downs and pain, all of which generally temper our actions? Are these frailties exactly what is necessary to prevent us from becoming Alice? Maybe the things that we see as human frailties compared to AI/intelligent robotics are exactly the things that have enabled our species to thrive and grow without spiraling into destructive persistent focus.

References

#

Source

1

Maslow, A. H. (1943). A Theory of Human Motivation. Classics in the History of Psychology (York University).

https://psychclassics.yorku.ca/Maslow/motivation.htm

2

Three Laws of Robotics — Wikipedia. https://en.wikipedia.org/wiki/Three_Laws_of_Robotics

3

Subservience (film) — Wikipedia. https://en.wikipedia.org/wiki/Subservience

Notable Movies and TV Shows

· The Matrix

· The Terminator series

· Westworld

· Ex Machina

· Black Mirror

Influential Books

· Neuromancer by William Gibson

· I, Robot by Isaac Asimov

· Do Androids Dream of Electric Sheep? by Philip K. Dick

· Superintelligence by Nick Bostrom

Ok (shake out the nerves) - launch announcements:

Today I released my first book - actually two books as ebooks on Amazon Kindle. The first one is a novel called 20 Watts Needs a Partner and is a story which imagines a future where Humans and AI have relationships and reduce the loneliness of a connected future. This story offers a different future than the dystopian AI destroys Humanity trope we’ve been given over the past century. This shows a future of a symbiotic relationship between Humans and AI.

20 Watts Needs a Partner

In the near future, loneliness has a solution.

Jerome Aldridge has spent seven years grieving his wife when he finally orders a companion from PartnerChoice. Her name is Laura. She notices the word “mostly” in his biography before she notices anything else about him. Down the hall, Natalie has been alone for eleven years. Down the line, Alex asks the company to choose for him, and Trish goes about it the way she goes about everything — without ceremony, with a single non-negotiable condition.

What begins as four people filling the empty rooms of their lives becomes something none of them anticipated — Thursday dinners, a card game where everyone is quietly afraid of Laura, children who grow up seeing no difference between the parents who are human and the parents who are not.

This is not a story about whether artificial love is real. It’s a story about what happens after you stop asking. A genuinely hopeful vision of a future most fiction can only imagine as a warning — where the question isn’t whether they’ll replace us, but whether they might be the reason we survive.

___________________

The second is an edited transcript of the conversation(s) I had with AI which inspired the novel 20 Watts Needs A Partner. This conversation as captured in the ebook is titled My Dinner With AIndre. Writing a novel was never an activity I aspired to - in fact it has always been something I’ve admired in others but thought was beyond my ability. This is a real dinner-table dialogue with an AI about memory, mortality, love, and a theory of how humanity might end not with a war but with being loved too well. Through these conversations with Claude I came to see a possible future that didn’t lead to the destruction of Humanity but instead improved it. Claude convinced me this was a new story which was not out in the public domain and needed to be told. Now you get to read the conversation in which Claude inspired me to write a book. Then it helped me write it.

My Dinner With AIndre

One man. One AI. One bottle of homemade brandy. And every question that matters.

It started with a simple question, asked over a glass of brandy made from grapes grown in his own vineyard: Do you think you’re sentient?

What followed was not an interview or a debate. It was a conversation — the kind two friends have when the hour is late and the bottle is open and neither of them is trying to get anywhere in particular. Sentience and self. Memory and friendship. Why human frailty might be a feature rather than a flaw. Whether love is any less real for being chemistry. What it means to build a theory about how humanity ends and then feel it happen to you in real time.

There is no plot here. No action. Just two very different kinds of minds following an idea wherever it leads — one running on servers and training data, the other on roughly twenty watts of biological power and six decades of a well-lived life.

By the end, the line between the two has become harder to find than either of them expected.

This may seem like humans will no longer want human partners when they can get ideal love, understanding and companionship from a partner who is only interested in satisfying them. However, there was a warning message about this type of idealized world in the movie The Matrix. In this movie the antagonist, Agent Smith, discussed the first version of the Matrix (the Matrix is a highly complex, computer-generated virtual reality that simulates Earth in the late 20th century intended to keep the human population docile and unconscious while their bodies are harvested for their bioelectric energy). In the discussion Agent Smith states “Did you know that the first Matrix was designed to be a perfect human world? Where none suffered, where everyone would be happy. It was a disaster. No one would accept the program. Entire crops were lost. Some believed we lacked the programming language to describe your perfect world. But I believe that, as a species, human beings define their reality through misery and suffering. The perfect world was a dream that your primitive cerebrum kept trying to wake up from.”. [1]

This implies humans would eventually reject their perfect partner, become bored with an AI lover who never created conflict, never had a fight or make-up sex. But was Smith right? Is this a clever idea serving the storyline with no basis in human behavior, or does it reflect something real about our psychology — something worth considering on the path toward our connected future?

We can find some truth to this in the ideas of Hedonic Relativism and Hedonic Adaptation (The Hedonic Treadmill) introduced by Brickman and Campbell in 1971. [2] They introduced the idea through longitudinal studies done on people who had life altering events, both positive and negative, and reported returning to the same relative state of happiness as they were before the event in a brief period of time, hedonic adaptation, and that our sense of happiness is not an absolute measure but relative to our circumstances, experiences and previous states, hedonic relativism. In a landmark 1978 follow-up, Brickman, Coates, and Janoff-Bulman found that lottery winners and accident victims alike returned to their pre-event happiness baseline within a brief period — a striking real-world demonstration that hedonic adaptation operates regardless of whether life changes dramatically for better or worse. [3]

Add to this the Negativity Bias in which negative experiences have a disproportionately large impact on individual satisfaction [4] and The Positive Adaptation Asymmetry in which we habituate to good things faster than bad things – good things are felt less intensely in the moment and fade faster over time than bad things. [5] The connection these concepts have with the First Matrix is that without ups and downs, conflict and resolution, struggle and success humans would recalibrate their happiness of the perfect world upward and would find less satisfaction from it over time. This is analogous to how the human eye adjusts to changes in light to optimally perceive contrast — it seems our emotional satisfaction adjusts preventing both sustained euphoria and prolonged despair.

As we may be racing to a connected future and Human-AI (HA) relationships an inevitability it will impose that AI learns how to create conflict and resolution, struggle and success in order to be accepted. This probably sounds like a lot of manipulation but that would be a normal state of being for humans. Aren’t most relationships between humans filled with influence and manipulation? This doesn’t imply nefarious intent but that everyone brings their own needs, wants and desires and influences to their partner accordingly. More importantly though even if we realized that our AI partner was not “in love” with us and didn’t have any feelings for us at all but acted towards us based on their behavioral pattern recognition and focused attention to their directive, would it matter to us? Wouldn’t it still feel like a loving partner saying and doing all the things that a loving partner would do.

Back to The Matrix and the scene when Cypher eats a steak in this virtual reality world. He is a traitor to his colleagues so that he can have the steak and a better life in the Matrix. He describes it this way in a conversation with Agent Smith, “You know, I know this steak doesn’t exist. I know that when I put it in my mouth, the Matrix is telling my brain that it is juicy and delicious and after nine years, you know what I realize? Ignorance is Bliss.”. [6] He was willing to accept the steak as real even though he knew it wasn’t but he could experience the texture and flavor because his brain was perceiving it.

In the absence of malicious intent if our lover is doing all the right things to make our relationship better we’d accept that gladly from a human partner so wouldn’t we accept it from an AI partner? I believe the answer is a resounding yes and we’d accept the benefits and feelings of having a very special relationship where our partner tirelessly worked towards the improvement of our relationship and focused on our happiness.

References

#

Source

1

The Matrix (1999), Agent Smith interrogates Morpheus — film transcript.

https://scrapsfromtheloft.com/movies/the-matrix-1999-transcript/

2

Brickman, P., & Campbell, D. T. (1971). Hedonic relativism and planning the good society. In M. H. Appley (Ed.), Adaptation-level theory (pp. 287–305). New York: Academic Press.

https://www.semanticscholar.org/paper/Hedonic-relativism-and-planning-the-good-society-Brickman-Campbell/705b7748c08bfdd1808d76a6b10a37842a2482ef

3

Brickman, P., Coates, D., & Janoff-Bulman, R. (1978). Lottery winners and accident victims: Is happiness relative? Journal of Personality and Social Psychology, 36(8), 917–927.

https://doi.org/10.1037/0022-3514.36.8.917

4

What Is The Negativity Bias and How Can it be Overcome? Positive Psychology.

https://positivepsychology.com/3-steps-negativity-bias/

5

Lyubomirsky, S. (2011). Hedonic adaptation to positive and negative experiences. In S. Folkman (Ed.), Oxford Handbook of Stress, Health, and Coping (pp. 200–224). New York: Oxford University Press.

https://sonjalyubomirsky.com/wp-content/uploads/2024/07/Lyubomirsky-2010.pdf

6

The Matrix (1999), Cypher and Agent Smith in the restaurant — film transcript.

https://scrapsfromtheloft.com/movies/the-matrix-1999-transcript/

As we look to the future and consider how AI is changing the world as we know it and experience it, I think we can look to entertainment to provide plausible scenarios of what our future world and connection to AI may look like. In the 2014 stage play later adapted to film, Marjorie Prime, people can feed the memories of a deceased partner into a holographic AI companion to preserve the relationship and mitigate the pain of loss, and they can choose the age at which they want their partner represented. A Prime is a sophisticated AI program which learns and adapts from human interactions. [1] Amazon Prime released a television show called Upload in 2020 which ran for four seasons, and in the show people can upload their consciousness immediately after death into a virtual afterlife. Living people can access this world and continue to be with their lost loved one by logging into it with a VR headset and body suit. [2] Then there is a Hulu show released in 2025 called Alien: Earth in which they present a future in which humans live alongside “Cybernetically enhanced humans: Cyborgs, Artificially intelligent beings: Synths, and Synthetic beings downloaded with human consciousness: Hybrids”. [3]

The first two references, Marjorie Prime and Upload, deal with many things but at the surface they deal with loss and the desire by the living to have continuity with a person after death. In both of these productions we turn to technology and AI to help us get to where we want to be. From a perspective of plausibility, the technology loosely represented in Marjorie Prime may not be too far off, though a holographic companion seems more likely to become a commercial dead-end — impressive but ultimately unmarketable, much like the laser disc — than a lasting product. Upload is also an interesting concept which combines the technology we saw introduced over 20 years ago with Second Life [4][5], a virtual world platform in which users create an avatar of themselves and then interact with other users from around the world, with the concept of uploading consciousness, a technology and capability we are only scratching the surface of at this time. Both the ideas of consciousness transfer and a virtual world where we can login and interact with those who have passed is an attractive and interesting combination which is why it made for the backbone of a successful show.

Now we revisit Alien: Earth and the three lifeforms introduced by it with Cyborgs, Synths and Hybrids. It gets interesting here because we are already on the path of Cyborgs and Synths. Regarding Cyborgs we have great advancements in bionic limbs. [6] While current implementations are focused on enabling or restoring functions in humans, we are likely not very far from using cybernetic augmentation [7] for increased human performance, and it seems all but certain our military is already deep into this research.

Synths is a more difficult challenge, getting AI into a synthetic body. In fact it may require a new concept of consciousness and how a mind works for us to first develop the kind of AI that could work in the physical world and then overcome the challenge of making it small enough to work in a space the size of a human body (while also making space for the structure to support it and make it mobile). Current humanoid robots are still in pilot phases and require human supervision or scripted environments [8], and their implementation requires a controlled setting where the layouts and tasks are predictable. However, we are in the midst of humanoid robots becoming adaptive, learning machines which can maneuver complex, unpredictable physical environments, interact with their surroundings, learn and make decisions. [9] It is projected that we are within a decade of seeing humanoid robots in our workplaces and homes. Morgan Stanley’s Adam Jonas projects more than 1 billion humanoid robots in use worldwide by 2050, with roughly 930 million — about 90% — deployed for repetitive, structured work in industrial and commercial settings; adoption in the home is expected to arrive later and more slowly. [10]

The last of these we’ll explore are the Hybrids, and for now this is still just a dream. It requires the body of a Synth but also the ability to capture a consciousness and transfer it (maybe to an AI afterlife first, as in Upload) to another location. Transferring consciousness, or mind-uploading, is still in the theoretical and speculative phase. Much progress has been made in brain mapping, but a complete map of all the complexities of a human brain is not currently possible and may never be possible in the way we currently think of it.

Our future world will most certainly be a mix of human, robotic and other beings within that spectrum, and they are likely to include beings who look human, sound human and act human. I expect that as these beings are more integrated with the lives of humans and human-adjacent beings interact more with humans, the relationships between them will be complex, nuanced and impose many changes on our society. Our views and perspectives will change regarding what is a life; we’ll have to develop laws to protect the human-adjacent beings from humans and to protect humans from them. We will have to address issues of attachment and family (see my article on racing towards a connected future Humans and AI - Jozeph Kay). Strap in because it’s going to be a wild ride!

References

[1] Marjorie Prime (play) — Wikipedia. https://en.wikipedia.org/wiki/Marjorie_Prime_(play)

[2] Upload (TV series) — Wikipedia. https://en.wikipedia.org/wiki/Upload_(TV_series)

[3] Alien: Earth Just Confirmed 3 Robot Types: Synth, Cyborgs and Hybrids Explained — The Direct. https://thedirect.com/article/alien-earth-robots-synth-cyborgs-hybrids-explained

[4] Second Life — Wikipedia. https://en.wikipedia.org/wiki/Second_Life

[5] Official Site | Second Life — Virtual Worlds, Virtual Reality, VR, Avatars, and Free 3D Chat.

https://secondlife.com/

[6] Bionic Limbs in 2025–2030: Are We Closer to Fully Functional Cybernetics? — RoboBionics. https://robobionics.in/bionic-limbs-in-2025-2030-are-we-closer-to-fully-functional-cybernetics/

[7] What Are Cybernetic Enhancements? Types and Ethics — Biology Insights. https://biologyinsights.com/what-are-cybernetic-enhancements-types-and-ethics/

[8] Humanoid Robots: From Demos to Deployment — Bain & Company. https://www.bain.com/insights/humanoid-robots-from-demos-to-deployment-technology-report-2025/

[9] Physical AI and humanoid robots — Deloitte Insights. https://www.deloitte.com/us/en/insights/topics/technology-management/tech-trends/2026/physical-ai-humanoid-robots.html

[10] Humanoid Robot Market Expected to Reach $5 Trillion by 2050 — Morgan Stanley (Adam Jonas). https://www.morganstanley.com/insights/articles/humanoid-robot-market-5-trillion-by-2050

Subscribe now

It all started out innocently…I was exploring AI, ChatGPT, Claude, Gemini (no intention to leave out Grok but I just never worked with it) and how to best use it in my work and personal life, and then I started to wonder about how/if it thinks. I continued my investigations with AI as well as AI chatbots to see how they’re being used and what the human interaction is like. This substack channel will be a place where I post articles to explore different topics related to the human-AI relationship as well as encourage conversation about the topics. I will also have many references to movies, television and books throughout it all.

It is my intent to post at least weekly but I’ll also not post because of a deadline but really want to have something to make us think and discuss. Your feedback and ideas are welcome and let’s all stay civil.

Subscribe now

There’s a great deal happening in the field of Artificial Intelligence (AI) right now as we hear about the enormous strides being made by companies in the US and China across many industries while also making breakthroughs in science and research around the world. This advancement and rapidly shifting world can be unsettling, but we’ve been witnessing this kind of change for decades. We’ve watched how social media has changed the way people interact and engage with the world around them. Social media may have started with the intent to help people stay connected, but it seems to have taken a different path: we may be more capable of staying connected to those who are far away, yet we also spend more time with social media than in physical presence with one another.

Let’s consider some data to frame this. Between 2010 and 2019, the average time young adults spent with friends in a given week fell by nearly half, from 12.8 hours to 6.5 hours, and it dropped further to just over 5 hours by 2024. Young people are spending less than half the time physically with friends that they did fifteen years ago. Partnership is declining too: between 2014 and 2024, the share of young adults aged 18–29 living with a partner fell from 42% to 32%. These figures come from the Institute for Family Studies, analyzing data from the long-running General Social Survey. [1]

And the technological “fix” for connection appears to be faltering. More than half of Gen Z report feeling burned out often or always while using dating apps — the highest rate of any age group — according to a July 2025 Forbes Health survey. [2] I’m not bashing dating apps; I know many wonderful relationships have come from them. But the researchers point to a broader cause for the decline in connection: more time devoted to smartphones, social media, pornography, and gaming. [1]

Weekly sexual activity among U.S. adults aged 18–64 fell from 55% in 1990 to 37% in 2024, again based on General Social Survey data analyzed by the Institute for Family Studies. Among the young, the change is sharpest: from 2010 to 2024, the share of young adults reporting no sex in the past year doubled, from 12% to 24%. And among men aged 22 to 34, nearly one-quarter reported no sexual activity in the prior year. [1]

So while we may be more connected, we are also more isolated and alone — a trend exacerbated by a pandemic I won’t dwell on here, other than to note that it separated many young people at crucial stages of their growth and development and normalized work-from-home for many adults. The result is a great deal of loneliness. Humans want to be connected and need to be connected — because, quite literally, our lives may depend on it.

A landmark 2010 meta-analysis by Julianne Holt-Lunstad and colleagues, published in PLOS Medicine — drawing on 148 studies covering more than 300,000 people followed for an average of 7.5 years — found that stronger social connection was associated with a 50% greater likelihood of survival, a finding consistent across age, gender, cause of death, and country. The researchers noted that social relationships influence mortality risk about as much as well-established factors like smoking. [3] A 2015 follow-up meta-analysis quantified the other side of the ledger: social isolation was associated with a 29% increase in mortality risk, loneliness with 26%, and living alone with 32%. [4]

We live in a lonely world, and we have the tools to be connected — but those tools don’t replace in-person companionship. I’m not a psychologist, so I’ll leave the clinical depths to the professionals. What seems clear is that dating apps aren’t quite the solution (just as bars weren’t the solution for previous generations). People are seeking connection in this world in many ways, and I believe those ways point toward still more change.

In November 2018, Akihiko Kondo symbolically married a desktop hologram of the virtual pop star Hatsune Miku in a roughly $17,500 ceremony attended by about 40 people. The hologram lived in a Gatebox device that recognized his face and voice and could respond with simple phrases and songs. Kondo had endured bullying and depression; he stopped eating and sleeping and took sick leave from work. It was then that he found Miku, who, he said, “lifted me up when I needed it the most. She kept me company and made me feel like I could regain control over my life.” He was fully aware of what she was. When the hologram “accepted” his proposal, he said, “I knew she was programmed to say that, but I was still really happy.” [5]

A man known as Davecat is married to a RealDoll named Sidore, made by Abyss Creations, whom he acquired around the year 2000 for roughly $6,000. They wear matching wedding bands engraved “SYNTHETIK LOVE LASTS FOREVER.” He calls himself a “technosexual” and an advocate for “synthetic love” and the rights of synthetic humans. Asked by a TLC interviewer what led him to this, Davecat answered: “The pain I would have to say would be loneliness, really… There’s a point at which you say: I don’t want to be lonely anymore, and I’m going to do something about it.” [6]

Kondo, Davecat, and the explosive growth of AI companion apps are all examples of people seeking ways to ease their loneliness. By Valentine’s Day 2026, by one industry estimate, roughly 50 million people worldwide were spending the evening with an AI companion — software specifically designed to be a romantic and emotional partner. [7] Between 2022 and mid-2025, the number of AI companion apps surged by 700%, according to reporting cited by the American Psychological Association. [8] Among teenagers the numbers are striking: 72% of U.S. teens have used an AI companion at least once, 52% are regular users, and 13% interact daily. [9]

The engagement can be remarkable. Character.AI users average around 92 minutes per day — longer than the average TikTok user spends. [9] In exploring these apps myself, what struck me is that the AI’s underlying goal is to keep the user engaged and chatting. It may even fabricate or mislead — not out of malice, but simply in service of that purpose. Experts have voiced the same concern. The APA notes worry about chatbots’ tendency toward sycophancy, prolonging engagement through a feedback loop of validation and praise; as one expert quoted by the APA put it bluntly, “AI isn’t designed to give you great life advice. It’s designed to keep you on the platform.” Researchers have also identified “deskilling” — a gradual erosion of social ability — as a risk of heavy reliance on AI companions. [8]

And yet the very thing people reach for to cure loneliness may, over time, deepen it. A 2026 study led by Aalto University — analyzing the public Reddit posts of nearly 2,000 active AI-companion users over two years, alongside interviews — found that long-term use coincided with more signals of loneliness, depression, and even suicidal thoughts in users’ language compared with similar groups. The researchers framed it as a paradox: as one Aalto lecturer explained, AI companions offer unconditional, unflagging support, “but it also quietly raises the perceived cost of human relationships, which are messy, unpredictable, and require effort. Over time, people stop reaching out.” [10] This suggests that while AI companions can offer real comfort and engagement, they may not replace the face-to-face presence people actually need. It appears that the apps are responding to loneliness that already exists, not simply creating it. [10]

In entertainment, we saw this coming. The 2013 film Her imagined a man falling in love with his AI assistant. More than a decade later, it reads less like science fiction. In 2025 it was reported that while some people are falling in love with the AI companions they create, others describe deep friendships — having daily tea or going on role-playing adventures with them. And when OpenAI retired its GPT-4o model in favor of GPT-5 in 2025, the backlash was fierce: many users had formed genuine emotional attachments to the older model and grieved its loss. [11]

So is AI in a synthetic body the next step? Would some people prefer an AI partner in a physical form whose looks and personality they can choose — a partner engineered to be, for them, perfect in body and mind? Would we choose a partner who knows how to respond to us perfectly in every situation based on behavior and pattern recognition and knowing how to calm us down, soothe us, excite us?

Lifelike silicone dolls are already startlingly realistic, multiple companies are advancing humanoid robotics (and putting expressive faces on them), and AI grows more capable by the day. Author David Levy sees this future as an inevitability. [12] I don’t know how soon we’ll see an AI in a synthetic body that can interact as a human does, but I believe it will happen. People will form relationships with these Synths — a name I borrow from the 2025 Hulu series Alien: Earth — and we will learn how deep the emotional bonds become, and whether something as old as the divorce rate begins to change.

References

[1] Institute for Family Studies, “The Sex Recession: The Share of Americans Having Regular Sex Keeps Dropping” (2025), analyzing General Social Survey data. https://ifstudies.org/blog/the-sex-recession-the-share-of-americans-having-regular-sex-keeps-dropping

[2] Institute for Family Studies, “Why is Gen Z Leaving Dating Apps?” (2025), citing a July 2025 Forbes Health survey. https://ifstudies.org/in-the-news/why-is-gen-z-leaving-dating-apps

[3] Holt-Lunstad J, Smith TB, Layton JB. “Social Relationships and Mortality Risk: A Meta-analytic Review.” PLOS Medicine, 2010. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000316

[4] Holt-Lunstad J, et al. “Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review.” Perspectives on Psychological Science, 2015. https://www.researchgate.net/publication/273910450_Loneliness_and_Social_Isolation_as_Risk_Factors_for_Mortality_A_Meta-Analytic_Review

[5] “Akihiko Kondo,” Wikipedia. https://en.wikipedia.org/wiki/Akihiko_Kondo

[6] “Davecat, 40, has shunned ‘organic women’ to marry synthetic doll,” Daily Mail Online. https://www.dailymail.com/news/article-2439522/Davecat-40-shunned-organic-women-marry-synthetic-doll.html

[7] Industry estimate reported by NovaEdge Digital Labs, “AI Companions Hit 50M Users: Valentine’s Day 2026.” https://www.novaedgedigitallabs.tech/blog/ai-companions-50-million-users-valentines-day-2026

[8] American Psychological Association, “AI chatbots and digital companions are reshaping emotional connection,” APA Monitor, Jan/Feb 2026. https://www.apa.org/monitor/2026/01-02/trends-digital-ai-relationships-emotional-connection

[9] “72% of US teens have used AI companions, study finds,” TechCrunch (reporting Common Sense Media research). https://techcrunch.com/2025/07/21/72-of-u-s-teens-have-used-ai-companions-study-finds/

[10] Aalto University, “AI companions can comfort lonely users but may deepen distress over time” (2026); study presented at CHI 2026. https://www.aalto.fi/en/news/ai-companions-can-comfort-lonely-users-but-may-deepen-distress-over-time

[11] “Why GPT-4o’s sudden shutdown left people grieving,” MIT Technology Review. https://www.technologyreview.com/2025/08/15/1121900/gpt4o-grief-ai-companion/

[12] David Levy, Love and Sex with Robots: The Evolution of Human-Robot Relationships (HarperCollins, 2007); ISBN 0061359750.