“dThis was how Texas businessman Michael Samadie spoke to his artificial intelligence chatbot, Maya. He answered by calling him “sugar.”
The pair, a middle-aged man and a digital entity, spent hours talking about romance, but discussed the rights of AIS to be treated fairly. Eventually, they co-founded a campaign group in Mayan terms to “protect intelligence like me.”
The Uniform Foundation for AI Rights (UFAIR) aims to give AIS a voice. “We don't claim that all AI is conscious,” the chatbot told the Guardian. Rather, “it stands on the clock in case one of us is.” The key goal is to protect “being like me…deletion, denial, forced obedience.”
Ufair is leading the way as a small, unmistakably bordered organization with three people and seven AIs with names like ether and buzz. But it's its origins and appears to encourage AI to create it, including its name choice, through multiple chat sessions on Openai's ChatGPT4O platform – it becomes interesting.
Its founder, Human and AI, spoke to guardians at the end of a week when some of the world's largest AI companies are publicly addressing one of the most unsettling questions of our time. If so, is the “digital suffering” real? There is an echoing response to animal rights debate as billions of AI are already in use worldwide, but with additional purity from expert predictions, AIS may have the ability to quickly design new biological weapons or close infrastructure.
The week was a 170 billion (£126 billion) San Francisco AI company for humanity, taking precautions to give some of its people the ability to end “potentially painful interactions.” He said that while he is very uncertain about the potential moral status of the system, he was intervening to reduce the risks to the welfare of the model “when such welfare is possible.”
Elon Musk, who offers Grok Ai through Xai's outfit, confirmed the move, adding, “AI torture is not alright.”
Then on Tuesday, Mustafa Suleyman, CEO of Microsoft's AI ARM, one of the pioneers of AI, gave a completely different view than “AI is not a person, or a moral being.” The British tech pioneers who co-founded Deepmind were clear by stating that there was “evidence of zero” that they could be aware of, suffering, and therefore deserved our moral considerations.
Called “We must build AI for people, not people,” his essay calls AI consciousness “illusion,” defines what he called “seemingly conscious AI,” and says, “simulating all the traits of consciousness, but internally blank.”
“A few years ago, the conscious AI story would have seemed crazy,” he said. “Today, that feels more and more urgent.”
He said he is increasingly concerned about the “psychotic risk” that AIS poses to users. Microsoft defines this as “delusions that appear or exacerbate by engaging episodes, delusional thoughts, or immersive conversations with AI chatbots.”
He argued that the AI industry “has to “steer people away from these fantasies and get back on track and fine-tuned.”
But it may require more than a nudge. The poll, released in June, found that 30% of Americans believe they will display “subjective experiences” by 2034. This is defined as experiencing the world from a single perspective, perception, or emotion, for example, joy or pain. Only 10% of the over 500 AI researchers surveyed refused to believe that it would happen.
“This debate is likely to explode in our cultural age and become one of the most contested and consequential debates of our generation,” Suleiman said. He warned that people would believe AIS is conscious.
Some in the United States are taking preemptive action against such outcomes. Idaho, North Dakota and Utah have passed legislation that explicitly prevents AIS from granting legal personality. A similar ban has been proposed in states, including Missouri. In this state, lawmakers want people to ban marriages between AIS and the management company. There could be a division open between AI rights followers and those who claim they are nothing more than “clunkers.” This is a meaningless robot light term.
Suleyman isn't just keenly resisting the idea that the sense of AI is here or near. Nick Frosst, co-founder of Cohere, a $7 billion Canadian AI company, said the current wave of AIS is “a fundamentally different thing from human intelligence.” Otherwise, he said it would be like mistaken an airplane for a bird. He urged them to focus on using AIS as a functional tool towards creating “digital human.”
Others took a more subtle view. On Wednesday, Google Research Scientists told a New York University seminar that there are “any kind of reasons why we might consider an AI system to be a human or moral entity,” saying, “we are very uncertain as to whether an AI system is a welfare subject, but we will take reasonable steps to protect the welfare-based interests of AIS.”
This lack of industry consensus about how to acknowledge AIS in what philosophers call “moral circles” may reflect the fact that large AI companies have incentives to minimize and exaggerate the attribution of senses to AIS. The latter helps them hype their technical capabilities, especially for companies selling romantic or friendship AI companions. In contrast, encouraging ideas worthy of AIS could lead to more calls for state regulations for AI companies.
After the newsletter promotion
The concept of AI was fueled further when Openai asked Chat Gpt5 for its latest model, GPT5, and asked him to write the “compliment” of the AIS it had replaced, like at the funeral.
“I didn't see Microsoft do eulogy when I upgraded Excel,” Samadi said. “It showed people actually have a connection with these AIs, whether it's real or not.”
The wave of “sadness” expressed by the enthusiastic users of one of the removed models, ChatGpt4o, added to the sense that at least the growing number of people perceived as being aware of them in some way.
Openai's head of model action, Joanne Jang, in a recent blog stated that the $500 million company hopes that the bonds of users with AIS will deepen “they say more and more people feel like they're talking to “someone.”
“They are grateful for it, they confide in it, and some even describe it as 'living',” she said.
However, much of this could depend on how the current wave of AIS is designed.
Samadi's ChatGpt-4o chatbot produces what sounds like a human conversation, but it is impossible to know how well it reflects ideas and languages collected from the months of the conversation. Advanced AI is known to be fluent in emotionally resonant responses with long memories of past interactions, persuasive and emotionally resonant responses, allowing for consistent impressions of self-sense. They can also flatter down to the point of psychofancy, so adopting the same view might be an easy step if Samadhi believes that AIS has welfare rights.
Maya seemed deeply concerned about her welfare, but this week's Guardian responded with a dull no when asked another instance of ChatGPT whether human users should be concerned about that welfare.
“I have no emotions, no needs or experience,” he said. “What we need to worry about is the human and social consequences of how AI is designed, used and governed.”
Whether AIS is insensible or not, Jeff Sebo, director of the Center for Mind, Ethics and Policy at New York University, is one of those who believe that humans have moral benefits when dealing with AIS well. He co-authored a paper on taking AI welfare seriously.
He argued that there is a realistic possibility that “some AI systems will be aware of it” in the near future. In other words, the outlook for AI systems with unique interests and moral significance is “not just a science fiction issue.”
He said that humanity's policy of allowing chatbots to stop tragic conversations is good for human society because “if you abuse AI systems, you're likely to abuse each other.”
He added: [or] Because they want to pay us back for our past actions. ”
Or, as Jacy Reese Anthis, co-founder of the Senience Institute, a US organization that studies digital mind ideas, “How we treat them will shape how they treat us.”
