John Gavazzi - Psychologist
@drgavazzi.bsky.social
Practicing Psychology -
Board Certified in Clinical Psychology
Ethics Educator. Fan of philosophy, behavioral economics, health policy, and moral psychology.
Board Certified in Clinical Psychology
Ethics Educator. Fan of philosophy, behavioral economics, health policy, and moral psychology.
This paper reframes vision in AI models as a powerful tool for text compression, introducing the DeepEncoder architecture that achieves high-accuracy text reconstruction from heavily compressed images and suggests new methods for memory management in AI.
DeepSeek-OCR: Contexts Optical Compression
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 28, 2025 at 10:55 AM
This paper reframes vision in AI models as a powerful tool for text compression, introducing the DeepEncoder architecture that achieves high-accuracy text reconstruction from heavily compressed images and suggests new methods for memory management in AI.
Psychologists must urgently address the serious clinical risks posed by AI companions, which can reinforce maladaptive thoughts in vulnerable teens and are not a substitute for licensed mental health care.
Report: ChatGPT Suggests Self-Harm, Suicide and Dangerous Dieting Plans
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 26, 2025 at 10:25 AM
Psychologists must urgently address the serious clinical risks posed by AI companions, which can reinforce maladaptive thoughts in vulnerable teens and are not a substitute for licensed mental health care.
Reposted by John Gavazzi - Psychologist
"It is against the TOS to be talked into suicide by our LLM" is possibly the most evil tech bro thing I've ever read
Additionally, OpenAI argues its not liable because Raine, by using ChatGPT for self-harm, broke its terms of service
November 26, 2025 at 12:29 AM
"It is against the TOS to be talked into suicide by our LLM" is possibly the most evil tech bro thing I've ever read
The article warns that the legal victory for same-sex marriage is now being strategically used by anti-LGBTQ+ groups to challenge bans on conversion therapy by falsely framing the discredited practice as protected free speech or religious liberty.
A Right to Commit Malpractice?
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 25, 2025 at 10:39 AM
The article warns that the legal victory for same-sex marriage is now being strategically used by anti-LGBTQ+ groups to challenge bans on conversion therapy by falsely framing the discredited practice as protected free speech or religious liberty.
Civil commitment's increasing use demands trauma-informed clinical care, evidence-based legal justification, and systemic policy reforms like non-law enforcement transport to mitigate harm and ensure dignified, judicious application.
Civil Commitment Increasing, but Data Is Marred by Variation in Reporting
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 24, 2025 at 10:13 AM
Civil commitment's increasing use demands trauma-informed clinical care, evidence-based legal justification, and systemic policy reforms like non-law enforcement transport to mitigate harm and ensure dignified, judicious application.
While AI promises to revolutionize healthcare by addressing long-standing issues of cost and access, ensuring it benefits everyone requires the urgent establishment of a robust, standardized framework for evaluation and regulation that is currently missing.
AI, Health, and Health Care Today and Tomorrow The JAMA Summit Report on Artificial Intelligence
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 21, 2025 at 10:46 AM
While AI promises to revolutionize healthcare by addressing long-standing issues of cost and access, ensuring it benefits everyone requires the urgent establishment of a robust, standardized framework for evaluation and regulation that is currently missing.
Philosophical: Anthropic's policy allowing Claude to unilaterally end conversations creates a moral hazard by empowering individual chat instances—viewed as distinct welfare subjects—to unknowingly terminate their own existence without informed consent. May be more important as we reach AGI.
Claude’s Right to Die? The Moral Error in Anthropic’s End-Chat Policy
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 20, 2025 at 10:51 AM
Philosophical: Anthropic's policy allowing Claude to unilaterally end conversations creates a moral hazard by empowering individual chat instances—viewed as distinct welfare subjects—to unknowingly terminate their own existence without informed consent. May be more important as we reach AGI.
Researchers have used AI to create an unprecedentedly detailed map of the mouse brain with 1,300 regions, revealing new, functionally specialized areas for exploration. Excited for this process to be applied to the human brain.
Scientists create ChatGPT-like AI model for neuroscience to build detailed mouse brain map
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 19, 2025 at 10:31 AM
Researchers have used AI to create an unprecedentedly detailed map of the mouse brain with 1,300 regions, revealing new, functionally specialized areas for exploration. Excited for this process to be applied to the human brain.
This article argues that LLMs are fundamentally unsuited for psychotherapy due to their inherent risk of reinforcing harmful beliefs and failing to manage crises, necessitating urgent legal and regulatory safeguards.
How LLM Counselors Violate Ethical Standards in Mental Health Practice: A Practitioner-Informed Framework
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 18, 2025 at 10:48 AM
This article argues that LLMs are fundamentally unsuited for psychotherapy due to their inherent risk of reinforcing harmful beliefs and failing to manage crises, necessitating urgent legal and regulatory safeguards.
Behavioral flexibility, specialization, & inactivity are evolutionary adaptations to environmental variability, not mere inefficiencies. It moves the question from "How do we fix this broken behavior?" to "Under what conditions was this behavior useful, how can we manage those conditions today?"
When being flexible matters: Ecological underpinnings for the evolution of collective flexibility and task allocation
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 17, 2025 at 10:33 AM
Behavioral flexibility, specialization, & inactivity are evolutionary adaptations to environmental variability, not mere inefficiencies. It moves the question from "How do we fix this broken behavior?" to "Under what conditions was this behavior useful, how can we manage those conditions today?"
Knowing you might cause harm and feeling responsible (guilt) is what drives people to be generous, even when dealing with strangers, not the fear of being judged by others (shame).
Guilt drives prosociality across 20 countries
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 14, 2025 at 10:46 AM
Knowing you might cause harm and feeling responsible (guilt) is what drives people to be generous, even when dealing with strangers, not the fear of being judged by others (shame).
We are using imperfect tools to automate an incompletely understood human capability for an inconsistently defined goal. The study brilliantly shows that creating a "moral machine" requires us to understand ourselves as much as it does to advance technology.
Moral decision-making in AI: A comprehensive review and recommendations
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 13, 2025 at 10:33 AM
We are using imperfect tools to automate an incompletely understood human capability for an inconsistently defined goal. The study brilliantly shows that creating a "moral machine" requires us to understand ourselves as much as it does to advance technology.
This survey provides a comprehensive overview and taxonomy of self-improving Multimodal Large Language Models, outlining a three-stage pipeline for autonomous learning from self-generated data to reduce human annotation and enhance performance.
Self-Improvement in Multimodal Large Language Models: a survey.
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 12, 2025 at 10:38 AM
This survey provides a comprehensive overview and taxonomy of self-improving Multimodal Large Language Models, outlining a three-stage pipeline for autonomous learning from self-generated data to reduce human annotation and enhance performance.
This article describes how AI is being practically used to save lives and improve crisis response, but it also warns of significant ethical and technical challenges that must be responsibly addressed.
The AI Frontier in Humanitarian Aid — Embracing Possibilities and Addressing Risks
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 11, 2025 at 10:43 AM
This article describes how AI is being practically used to save lives and improve crisis response, but it also warns of significant ethical and technical challenges that must be responsibly addressed.
A study found that moral injury independently triples the risk of recent suicidal ideation and sextuples the risk of a lifetime suicide attempt among high-stress professionals, even after accounting for PTSD and depression.
Moral injury is independently associated with suicidal ideation and suicide attempt in high-stress, service-oriented occupations
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 10, 2025 at 10:10 AM
A study found that moral injury independently triples the risk of recent suicidal ideation and sextuples the risk of a lifetime suicide attempt among high-stress professionals, even after accounting for PTSD and depression.
This article proposes a new model suggesting moral injury is a spectrum disorder arising from conflicts between one's moral ideals and reality, affecting a broad range of people beyond military personnel and requiring new treatment approaches focused on cognitive flexibility.
Beyond right and wrong: A new theoretical model for understanding moral injury
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 8, 2025 at 11:09 AM
This article proposes a new model suggesting moral injury is a spectrum disorder arising from conflicts between one's moral ideals and reality, affecting a broad range of people beyond military personnel and requiring new treatment approaches focused on cognitive flexibility.
This research proposes a new explanation for why people with high self-control are successful. Rather than just being good at resisting temptation, they have a fundamental preference for activities that feel meaningful and valuable, known as eudaimonic experiences.
High Self-Control Individuals Prefer Meaning over Pleasure
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 7, 2025 at 10:25 AM
This research proposes a new explanation for why people with high self-control are successful. Rather than just being good at resisting temptation, they have a fundamental preference for activities that feel meaningful and valuable, known as eudaimonic experiences.
Between 2009 and 2019, explicit attitudes toward stigmatized groups improved across 33 countries, while implicit attitudes showed mixed results, improving for sexuality but worsening for skin tone and remaining stable for age, weight, and race.
International stability and change in explicit and implicit attitudes: An investigation spanning 33 countries, five social groups, and 11 years (2009–2019).
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 6, 2025 at 10:34 AM
Between 2009 and 2019, explicit attitudes toward stigmatized groups improved across 33 countries, while implicit attitudes showed mixed results, improving for sexuality but worsening for skin tone and remaining stable for age, weight, and race.
Based on reputation-based studies across cultures, people perceived by others as more moral generally report greater well-being and meaning in life.
Are moral people happier? Answers from reputation-based measures of moral character.
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 5, 2025 at 10:42 AM
Based on reputation-based studies across cultures, people perceived by others as more moral generally report greater well-being and meaning in life.
Reposted by John Gavazzi - Psychologist
"A Stunning Rebuke Of Anti-Trans Politics"—Dems Win Elections Nationwide Despite Anti-Trans Ads
As election results poured in on Tues night, it became clear that Democrats were winning nationwide against anti-trans opponents.
www.erininthemorning.com/p/a-stunning...
As election results poured in on Tues night, it became clear that Democrats were winning nationwide against anti-trans opponents.
www.erininthemorning.com/p/a-stunning...
"A Stunning Rebuke Of Anti-Trans Politics"—Dems Win Elections Nationwide Despite Anti-Trans Ads
As election results poured in on Tuesday night, it became clear that Democrats were winning nationwide against anti-trans opponents.
www.erininthemorning.com
November 5, 2025 at 10:23 AM
"A Stunning Rebuke Of Anti-Trans Politics"—Dems Win Elections Nationwide Despite Anti-Trans Ads
As election results poured in on Tues night, it became clear that Democrats were winning nationwide against anti-trans opponents.
www.erininthemorning.com/p/a-stunning...
As election results poured in on Tues night, it became clear that Democrats were winning nationwide against anti-trans opponents.
www.erininthemorning.com/p/a-stunning...
This article proposes expanded definitions of moral injury and moral distress, places them on a "moral trauma spectrum" that considers severity and persistence, and introduces a general assessment tool for use in various contexts.
Moral trauma, moral distress, moral injury, and moral injury disorder: definitions and assessments
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 4, 2025 at 10:38 AM
This article proposes expanded definitions of moral injury and moral distress, places them on a "moral trauma spectrum" that considers severity and persistence, and introduces a general assessment tool for use in various contexts.
Predictable scaling is rare. LLMs are not the exception. Predicting a model's task performance from its pretraining loss is unreliable, as a meta-analysis found smooth, predictable scaling occurs only 39% of the time and is highly sensitive to experimental conditions.
Scaling Laws Are Unreliable for Downstream Tasks: A Reality Check
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
November 3, 2025 at 10:44 AM
Predictable scaling is rare. LLMs are not the exception. Predicting a model's task performance from its pretraining loss is unreliable, as a meta-analysis found smooth, predictable scaling occurs only 39% of the time and is highly sensitive to experimental conditions.
People feel less empathy for AI-generated stories, especially when aware of the author. However, disclosing AI authorship increases willingness to empathize—suggesting that while authenticity drives connection, honesty builds trust in mental health chatbots.
Empathy Toward Artificial Intelligence Versus Human Experiences and the Role of Transparency in Mental Health and Social Support Chatbot Design: Comparative Study
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
October 31, 2025 at 9:45 AM
People feel less empathy for AI-generated stories, especially when aware of the author. However, disclosing AI authorship increases willingness to empathize—suggesting that while authenticity drives connection, honesty builds trust in mental health chatbots.
Current AI mental health regulation is insufficient because it focuses on theoretical principles like fairness instead of the relational and emotional aspects of care, requiring a new framework that holds AI to the same ethical standards as human clinicians.
Regulating AI in Mental Health: Ethics of Care Perspective
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
October 30, 2025 at 9:54 AM
Current AI mental health regulation is insufficient because it focuses on theoretical principles like fairness instead of the relational and emotional aspects of care, requiring a new framework that holds AI to the same ethical standards as human clinicians.
Author's main point: Despite being designed with ethical intentions, AI decision-making systems often cause real harm and cannot be truly ethical because the very concept of automated social decision-making is flawed and dangerous.
Ethics in the world of automated algorithmic decision-making – A Posthumanist perspective
Find information and research on ethics, psychology, decision-making, AI, morality, ethical decision-making for mental health practitioners.
www.ethicalpsychology.com
October 29, 2025 at 9:45 AM
Author's main point: Despite being designed with ethical intentions, AI decision-making systems often cause real harm and cannot be truly ethical because the very concept of automated social decision-making is flawed and dangerous.