Daniel Did Not Need a Chatbot: He Needed a System That Saw Him

During the week Daniel died, twenty-six emails were sent to psychiatrists seeking help. Only two received replies, and both arrived after he was gone.

That fact sits at the center of my work on artificial intelligence and mental health. Young people increasingly turn to technology because it is immediate, private, and available when human care is not. Their choice does not prove that a chatbot is an adequate therapist. It reveals how frequently the existing system fails to answer.

Daniel did not need simulated empathy. He needed coordinated human care capable of seeing his complete medical story.

His information was fragmented across psychiatric treatment, substance-use care, emergency medicine, physical health, medications, trauma, and family communications. Each professional saw a portion. No one appeared to hold the complete picture. Symptoms were treated, labels accumulated, and the burden of connecting everything fell on a young man in crisis and his family.

My chapter in How to Thrive in the AI Transformation examines what responsible technology might contribute. AI may eventually help qualified clinicians organize longitudinal records, identify medication interactions, detect meaningful changes, flag missed follow-up, and reveal patterns hidden across disconnected systems. It could help teams ask better questions sooner.

But possibility is not proof.

Artificial intelligence cannot independently diagnose the full complexity of a person, predict suicide with certainty, replace clinical judgment, or assume responsibility for someone in crisis. Systems can hallucinate, reflect bias, misunderstand culture, miss medical context, and create false confidence. A compassionate tone is not the same as safe treatment.

Responsible innovation therefore requires boundaries. Mental-health technology should be clinically validated, transparent about its limitations, designed with trauma-informed principles, and governed by meaningful human oversight. People must know when they are interacting with a machine. Their most intimate data must not become a commercial product.

Crisis pathways must connect users to trained human beings. Technology should never imitate a person who died or exploit bereavement by manufacturing artificial contact.

The question is not whether AI is good or bad. The question is what problem we are asking it to solve—and who remains accountable when it fails.

That is the standard I carry: innovation must increase human attention, accountability, and timely access to care—not merely produce another convenient digital interface.

At the Daniel Calazans Legacy, our interest in AI begins with the whole person and the whole family. A young adult’s safety cannot be separated from medical history, substance exposure, sleep, trauma, relationships, financial pressure, or the people trying to obtain help. Nor does a family’s need for care end when a loved one dies.

Technology should strengthen relationships among patients, families, and professionals, not replace them. It should make fragmented systems more attentive, not allow institutions to withdraw further behind automation.

Daniel loved technology, robotics, and difficult questions. Honoring him requires both imagination and skepticism. The future of mental-health AI should not be built around machines that sound human. It should be built around systems that help humans see, answer, and care before silence becomes irreversible.