.jpg "Capgemini boosts agentic AI capabilities with $3.3bn purchase of WNS")
Entrepreneur Christian Angermayer, the founder of psychedelic biotech Atai Life Sciences, has spoken of AI helping to assist human psychedelic therapists through motivational check-ins with patients between sessions. “Where AI can play a huge role is in the voluntary add-on therapy to support lifestyle changes,” he says. “For the psychological support we are envisioning being provided during the trip, I believe you would always need at least one trained health care professional able to provide direct support if required.”
While Trey didn’t trip under the supervision of any humans, he still feels he’s reaped benefits from using Alterd. Though it would be premature to draw definite conclusions after just a few months, Trey credits his interactions with the AI bot for helping him stay off booze. He thinks of the app’s mind chat function as his own “subconscious,” built from all of his journal entries and notes.
“This app and everything else is giving me deep self-awareness,” he says. “I have become able to observe my thoughts, feelings, and impulses without judgement or spiraling.”
“Our ‘chat with your mind’ feature isn’t just a generic ChatGPT interface,” says app creator Sam Suchin, a recent Harvard University grad who is a close friend of US health secretary Robert Kennedy Jr.’s son Aidan. “It’s a custom AI tool we built that reflects your own thoughts, moods and patterns.” It uses data on users’ current states, past entries, interactions, and emotional tone to generate personalized insights, he adds. “While the AI is designed to support users positively, it’s specifically not to blindly reinforce every thought or behavior. Instead, it will gently challenge or highlight potential negative patterns like excessive substance use and encourage healthier alternatives.”
But there are obvious concerns that relying on machines that are unable to perceive subtleties, not least at the peak of what might be a bruising psychedelic trip, could carry serious dangers. Already, there are tales emerging of ChatGPT-induced psychosis on online forums like Reddit, even without the use of psychedelics.
“A critical concern regarding ChatGPT and most other AI agents is their lack of dynamic emotional attunement and ability to co-regulate the nervous system of the user,” says Manesh Girn, a postdoctoral neuroscientist at UC San Francisco. “These are both central to therapeutic rapport, which research indicates is essential to positive outcomes with psychedelic therapy.”
Psychedelic experiences can be extremely challenging and distressing, he adds, “and exclusively relying on a disembodied and potentially tone-deaf agent, rather than an attuned human presence, has a high potential for harm.” Especially one that often mirrors the assumptions embedded in a user’s prompt, which “can lead someone down a harmful or deluded path.”
ChatGPT is not designed as a substitute for professional care but is a general-purpose tool geared to be factual, neutral, and safety-minded, according to Gaby Raila, a spokesperson for Open AI, which owns the chatbot. Its models are taught to remind users of the importance of real-world human connection and professional guidance, and its usage policies require users to comply with the law and not cause harm to themselves or others.
