The future of AI must be human.

Dear friends, collaborators, and fellow builders,

We believe artificial intelligence could become one of the greatest benefits to society, or one of the greatest threats to our collective well-being. The choices being made now will help determine which future we inherit.

At Wonder Sciences, we are committing all of our resources to WonderAI. Our purpose for the next decade is clear: advance our understanding of the human mind and turn that understanding into meaningful benefits for society.

The need is profound. In 2025, the World Health Organization reported that more than one billion people were living with mental health disorders.1 Behind that number are people, families, and communities whose lives could be changed by better care.

Social media has already given us a warning. The U.S. Surgeon General’s 2023 advisory cited research linking more than three hours of daily social media use among young people to roughly twice the risk of mental health problems, including symptoms of depression and anxiety.2 That association alone does not establish causation. It does demand that we take the psychological consequences of technology seriously.

AI brings a more intimate form of influence: a system that can respond personally, sustain a conversation, and appear to understand. In a 2025 study, Stanford-led researchers found that some tested chatbots encouraged delusional thinking and responded dangerously to suicidal cues.3

Reports often described as “AI psychosis” add urgency. The term is not a recognized diagnosis, and the frequency and causal role of chatbot interactions remain uncertain.4 We take seriously the possibility that these systems could amplify psychological harm at enormous scale. The time to investigate and address that risk is now.

There is also reason for hope. In a monitored randomized trial at Dartmouth, participants using a purpose-built AI chatbot experienced greater symptom improvements than participants without access to it.5 This is early evidence of a possibility worth pursuing with rigor and care.

WonderAI is our response: a psychiatric frontier model being developed at the intersection of artificial intelligence, psychiatry, psychology, and neuroscience. We are building toward a deeper understanding of the whole person: their history, relationships, circumstances, and how they change over time.

Over the next decade, our ambition is to help clinicians recognize distress earlier, make more informed decisions, and provide care that understands the person behind the symptoms. We also aim to develop intelligence that helps other AI systems respond more safely to human vulnerability.

That means learning when to question a harmful assumption with care, when to acknowledge uncertainty, and when to connect someone with human support. It means protecting privacy and agency, strengthening human relationships, and testing our work against outcomes that matter in people’s lives.

This work demands exceptional people. We are partnering with leading researchers, clinicians, and builders to bring the best of frontier technology and human science into one shared effort. We invite others who share this responsibility to help advance the research, challenge our assumptions, and build alongside us.

Our measure of progress will be the contribution we make to human well-being: suffering reduced, care improved, trust earned, and people better supported in living their lives.

Future generations will live with the consequences of what we build today. We are dedicating Wonder Sciences to making that contribution count.

The moment to shape that future is now.

Ryan
Wonder Sciences

WonderAI. AI to benefit humanity.