1. Algorithm Aversion

Question: “If an AI therapy app gave harmful advice once, should patients ever trust it again, or is one mistake enough to end its credibility?”

Dr. Jeune: “I am all for progress, but human machines are very tricky. Psychological medicine is not comparable to giving directions/instructions on repairing a fault in a mechanical engine. Harmful advice can end a life at worst and harm several lives at best. However, the reality is that when a person is distressed they will seek out any support that may soothe them. Although in my opinion giving harmful therapeutic advice would lose my confidence, there are many patients who may have a different attitude and worldview that would not be as dismissive as I am.”

Dr. Jeune’s distinction between professional standards and patient desperation sets the stage for how emotional availability governs trust in technology.

2. Automation Bias

Question: “Why do you think some professionals follow AI advice even when their own judgment tells them it might be wrong?”

Dr. Jeune: “All self-respecting professionals work between minimal standards and maximum standards. As human machines, they are affected by their mood, emotional availability and knowledge base. It is my view that professionals that are working at their minimal standard would be more likely to take AI advice even if their own experience and knowledge base contradicts that advice. When a professional is working at their maximum standard they will trust their judgement over anyone else’s including AI.”

This internal fluctuation in professional performance explains why reliance on AI is often a symptom of human fatigue rather than a lack of competence.

3. Algorithms Codifying Injustice

Question: “If AI learns from biased human data, can it ever really be fair, or will it always repeat society’s prejudices?”

Dr. Jeune: “If data is biased then it cannot be fair can it? There is no practical reason why AI cannot give samples of a range of views and explanations. When you train as a psychologist, you learn a range of personality theories and apply what fits the issue best. There is no reason why a patient should not ask the same question in slightly different ways to see if there is integrity in the information—consistent, repeated messages—which would reduce the risk of bias.”

By advocating for multiple perspectives rather than a singular truth, Dr. Jeune suggests a path toward bias reduction through integrative, patient-centered inquiry.

4. Illusion of Explainability

Question: “In psychology, would patients feel reassured by a simple AI explanation of their condition, even if it wasn’t entirely accurate?”

Dr. Jeune: “Patients are not homogenous. They all have different attitudes, personalities, moods, intellect, communication styles, and comprehension ability. Some may want a ‘quick fix’ and therefore are prepared to accept any explanation that is expedient or reduces their angst the quickest. So if they just want a generic explanation in that moment, it will be sufficient, but if they want something more permanent or enduring as a solution, then they will not be satisfied.”

This variance in user depth-of-need highlights the fundamental impossibility of a one-size-fits-all AI solution for emotional health.

5. AI Taking Jobs

Question: “Could AI ever replace the human empathy at the heart of therapy, or will psychology always need the human touch?”

Dr. Jeune: “In my opinion, AI cannot replicate the therapeutic encounter. The responses and reactions that pass between a therapist and the help-seeker can be palpable. However, just because something cannot tell you everything does not mean that it cannot tell you anything. For those who do not value human contact, it may be satisfying to them. But for those who form the majority of humankind, those thinking, feeling, caring, sociable people, AI will be quite insufficient.”

Concluding with a balanced view, Dr. Jeune reminds us that while AI may serve a specific niche, it remains a shadow of the essential, palpable human encounter.