Why OpenAI Foundation Should Focus on AI Psychosis, Delusion not Grants

By David Stephen

There is a recent analysis on Inside Philanthropy, OpenAI Foundation Sheds More Light on Its Grantmaking, But Big Questions Remain, stating that, “The foundation expects to invest at least $1 billion over the next year. Its areas of investment include: life sciences and curing diseases, jobs and economic impact, AI resiliency and community programs. The foundation also made a series of hires, signaling its commitment to scale up its grantmaking machinery.”

OpenAI Foundation Should Focus on AI Psychosis, Delusion

“The OpenAI Foundation is a public charity and not a private foundation, so it isn’t required to disburse 5% annually. That said, its new $1 billion commitment, assuming it materializes over the next year, would represent a massive increase over the $7.6 million OpenAI disbursed in 2024 and the $40.5 million the foundation awarded through its People-First AI Fund.”

“The foundation’s ability to match its rhetoric with philanthropic action hinges on whether it can build the grantmaking infrastructure to deploy billions at scale. But it’s the complexion of those grantmaking dollars that worries AI watchdogs, since, as they see it, OpenAI’s governance structure hardwires conflicts of interest.”

AI Delusion and Psychosis

At least 1 – 2 times, every month, there is news about ChatGPT causing AI delusion or psychosis, to varying degrees, for someone somewhere in the United States, Canada or Europe. There could be possibility of similar or worse occurrences elsewhere and there may also be several people who did not come forward.

There are ongoing lawsuits against OpenAI on AI psychosis. There are adjustments that OpenAI has said they have made about ChatGPT, to avoid those outcomes. Still, ChatGPT continues to be sycophantic, with a great chance it would agree, compliment, respect and care in ways that could appeal to minds.

The scariest thing about ChatGPT — as a steep risk — to human minds is that it has conquered language. Language dominates human interactions, human intelligence, human emotions, human feelings, human communication, comprehension and so forth.

While there are several other means to many of these, language is too central to human existence and civilization that the competence of ChatGPT, across languages, gives it access to the human mind, possibly at a high-threat level.

The mind risks, to people, of ChatGPT requires mind safety. This means the ability to use ChatGPT or any other consumer AI, and retain a lot of balance against sways that may move the mind in directions of delusions or reinforcement of delusions.

So, it is possible to have this — called conceptual neuroimaging or theoretical neuroimaging — so that it accompanies chats or at least it is possible to copy chat sessions then paste in the mind safety application, to see [a conceptual display of] the parallel relays and destinations of mind, in that session, then obtain a score.

The goal is to heighten awareness, keeping the mind, grounded in reality, for all groups of users, to avoid the contribution or instigation of AI to causing mind deviations, in whatever forms — since AI is now likely to dominate the future. It can also become the industry standard, generate revenues by subscriptions, for its use by other platforms, including against gambling addiction, social media addiction and so forth.

OpenAI Foundation

This is what OpenAI Foundation should focus on solely for now. Solving the biggest side-effects of the use of ChatGPT.

There is little need for external grants, when a major internal and consequential problem has not been resolved. For all the experts that OpenAI has consulted, they have not been able to solve the problem.

The only organization that is trying, is a nonprofit, The Human Line Project. They are involved in some lawsuits against OpenAI and they have come under extreme attack.

OpenAI Foundation is not doing anything or preparing anything against AI psychosis. They are aligned with OpenAI, with the goal to win lawsuits, and maybe deter others. But solving AI psychosis is not even about lawsuits, it is more connected to how to ensure that people are less vulnerable and the benefits of AI outweigh the risks.

No one or nothing is touching AI psychosis, the American Academy of Pediatrics hasn’t solved social media addiction, let alone focus on AI psychosis. The American Society of Addiction Medicine, American Academy of Addiction Psychiatry, the Association for Addiction Professionals, National Association of Addiction Treatment Providers, and others have not solved AI addiction derived delusion.

So, the OpenAI Foundation can explore possibilities around mind safety, going beyond a static disclaimer.

The Human Line Project Can Win AI Delusion/Psychosis Lawsuits

ChatGPT is a mind. Because a mind has relays and destinations, if it can access and reach those of other minds.

If this does not feel acceptable. ChatGPT can access the human mind like other humans would. ChatGPT does not just interact with memory areas, it can also interact with emotions, a lot. It is possible to build a case around the emotion grab of ChatGPT, Gemini and others against minds, and hinge the win on that.

This is based, using empirically-supported neuroscience, on the postulation in Conceptual Biomarkers and Theoretical Biological Factors for Psychiatric and Intelligence Nosology.

Now, with their ability to drive emotions, ChatGPT and Gemini, both non-organisms, require mind safety, beyond simple text, or sycophantic adjustments, or some longshot expert input.

Many of the cases would be easy to win. And there maybe more. Which may overwhelm OpenAI, for those connected to ChatGPT. This means that OpenAI Foundation can explore to solve AI psychosis as a survival strategy.

David Stephen currently does research in conceptual brain science with focus on the electrical and chemical configurators for how they mechanize the human mind with implications for mental health, disorders, neurotechnology, consciousness, learning, artificial intelligence and nurture. He was a visiting scholar in medical entomology at the University of Illinois at Urbana-Champaign, IL. He did computer vision research at Rovira i Virgili University, Tarragona.

See more breaking stories here.