Automated Digital Safety Planning Interventions for Young Adults: Qualitative Study Using Online Co-design Methods (Preprint) DOI
Jonah Meyerhoff, Sarah A Popowski, Tanvi Lakhtakia

и другие.

Опубликована: Дек. 3, 2024

BACKGROUND Young adults in the United States are experiencing accelerating rates of suicidal thoughts and behaviors but have lowest formal mental health care. Digital suicide prevention interventions potential to increase access care by circumventing attitudinal structural barriers that prevent These tools should be designed collaboration with young who lived experience suicide-related optimize acceptability use. OBJECTIVE This study aims identify needs, preferences, features for an automated SMS text messaging–based safety planning service support self-management among adults. METHODS We enrolled 30 (age 18-24 years) recent participate asynchronous remote focus groups via online private forum. Participants responded researcher-posted prompts were encouraged reply fellow participants—creating a threaded digital conversation. Researcher-posted centered on participants’ experiences thought behavior-related coping, planning, technologies behavior self-management. Focus group transcripts analyzed using thematic analysis extract key feature considerations tool. RESULTS adult participants indicated message–based intervention must meet their needs 2 ways. First, empowering them manage symptoms own acquiring effective coping skills. Second, leveraging adults’ existing social connections. also shared 3 technological intervention: (1) transparency about how functions, kinds actions it does not take, limits confidentiality, role human oversight within program; (2) strong privacy practices—data security around content data created would maintained used was extremely important given sensitive nature data; (3) usability, convenience, accessibility particularly participants—this includes having approachable engaging message tone, customizable delivery options (eg, length, number, focus), straightforward menu navigation. highlighted specific could core skill acquisition self-tracking, idea generation, reminders). CONCLUSIONS Engaging design process tool revealed critical addressed if is effectively expand evidence-based reach people at risk behaviors. Specifically, building skillfulness cope crises, deepening interpersonal connections, system transparency, privacy.

Язык: Английский

Rethinking technology innovation for mental health: framework for multi-sectoral collaboration DOI
Jina Suh, Sachin R. Pendse, Robert Lewis

и другие.

Nature Mental Health, Год журнала: 2024, Номер 2(5), С. 478 - 488

Опубликована: Май 3, 2024

Язык: Английский

Процитировано

3
Evaluating the Cybersecurity Robustness of Commercial LLMs against Adversarial Prompts: A PromptBench Analysis DOI Creative Commons

Takeshi Goto,

Kensuke Ono, Akira Morita

и другие.

Опубликована: Март 6, 2024

This study presents a comprehensive evaluation of the cybersecurity robustness five leading Large Language Models (LLMs)-ChatGPT-4, Google Gemini, Anthropic Claude, Meta Llama, and Mistral 8x7B-against adversarial prompts using PromptBench benchmark. Through dual approach quantitative qualitative analysis, research explores each model's performance, resilience, vulnerabilities. Quantitative metrics such as accuracy, precision, recall, F1 scores offer statistical comparison across models, while insights reveal distinct patterns response susceptibility to various strategies. The findings highlight significant variations in model robustness, underlining importance complex enhancing LLM security. not only sheds light on current limitations but also emphasizes need for advancing methodologies development practices mitigate potential threats ensure safe deployment LLMs sensitive critical applications.

Язык: Английский

Процитировано

2
Large Language Model Agents for Improving Engagement with Behavior Change Interventions: Application to Digital Mindfulness DOI
Harsh Kumar, Suhyeon Yoo, Angela Zavaleta Bernuy

и другие.

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

2
Making Moral Decisions With Artificial Agents As Advisors. AnfNIRS Study DOI Creative Commons
Eve Fabre,

Damien Mouratille,

Vincent Bonnemains

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Март 11, 2024

Abstract Artificial Intelligence (AI) is on the verge of impacting every domain our lives. It increasingly being used as an advisor to assist in making decisions. The present study aimed at investigating influence moral arguments provided by AI-advisors (i.e., decision aid tool) human decision-making and associated neural correlates. Participants were presented with sacrificial dilemmas had make decisions either themselves baseline run) or that utilitarian deontological AI-advised run), while their brain activity was measured using f NIRS device. Overall, significantly influenced participants. Longer response times a decrease right dorsolateral prefrontal cortex observed than arguments. Being machines appears have led decreased appraisal affective dilemmas. This resulted reduced level utilitarianism, supposedly attempt avoid behaving less cold-blooded way preserve (self-)image. Taken together, these results suggest motivational power can voluntary up- down- regulation processes along decision-making.

Язык: Английский

Процитировано

1
Designing a Large Language Model-Based Coaching Intervention for Lifestyle Behavior Change DOI
Sophia Meywirth

Lecture notes in computer science, Год журнала: 2024, Номер unknown, С. 81 - 94

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

1
Enhancing ethical codes with artificial intelligence governance – a growing necessity for the adoption of generative AI in counselling DOI
Pei Boon Ooi, G.G. Wilkinson

British Journal of Guidance and Counselling, Год журнала: 2024, Номер unknown, С. 1 - 15

Опубликована: Июль 8, 2024

The advent of generative Artificial Intelligence (AI) systems, such as large language model chatbots, is likely to have a significant impact in psychotherapy and counselling the future. In this paper we consider current state AI evolution field. We examine ethical codes practice for four countries different parts world, namely UK, USA, Australia Malaysia, identify aspects these that will need enhancement reflect good governance. Using Model Governance Framework an example, identified how key elements framework relate core codes, pointer be enhanced if systems are adopted by profession.

Язык: Английский

Процитировано

1
Chatbot-Based Interventions for Mental Health Support DOI Open Access
Yiyi Wang, Norman A. S. Farb

Опубликована: Авг. 29, 2024

Objective: Mental health concerns are rising, particularly among post-secondary students, who may lack access to traditional therapeutic resources due barriers like long wait times and high costs. To help address these challenges, we explored the potential of large language model-based chatbots for supporting mental wellbeing in student populations.Methods: We conducted two studies, lasting one week four weeks, examine effectiveness chatbot interventions over different durations. Both studies compared interventions—one mindfulness-focused value-focused—against an active check-in-only control condition. The primary outcome measure was improvement through mindfulness-to-meaning (MM) pathway, a process which enhanced decentering, ability see one’s experience from wider perspective, leads improved positive reappraisal, find constructive empowered interpretations experience.Results: All conditions showed evidence stress reduction. However, group, both intervention styles at durations resulted via MM pathway. This effect primarily driven by significant improvements decentering. For longer duration only, also observed reappraisal. Conclusions: These results emphasize chatbot-based support development regulatory skills leveraging pathway enhance health. Educational institutions providers might consider integrating such tools into scalable accessible systems, addressing broader more diverse audience while promoting sustained development.

Язык: Английский

Процитировано

1
Digitalisierung und maschinelles Lernen in der Psychotherapieforschung und Praxis – Potentiale und Probleme DOI
Miriam I. Hehlmann, Wolfgang Lutz

PPmP - Psychotherapie · Psychosomatik · Medizinische Psychologie, Год журнала: 2023, Номер 73(09/10), С. 367 - 369

Опубликована: Окт. 1, 2023

Im Zuge des weltweiten Anstieges der Bedeutung von psychischen Störungen 1, werden frühzeitige Interventionen und wirksame psychotherapeutische Behandlungen für ein funktionierendes Gesundheitssystem immer wichtiger. Der aktuelle Stand Psychotherapieforschung zeigt jedoch, dass nicht alle Patient:innen gleichermaßen Psychotherapie profitieren, sondern die meisten (70–80%) zwar deutliche Verbesserung zeigen, während andere nur geringe oder keine Fortschritte erzielen sogar Verschlechterungen erfahren 2. Dies impliziert eine stärkere Berücksichtigung individuellen Unterschieden deren Therapieverlauf in Psychotherapieforschung, sowie Refokussierung auf ungünstige Therapieverläufe Abkehr Frage nach durchschnittlichen zwischen den verschiedenen Therapieverfahren Therapieschulen.

Процитировано

1
Mixed Method Feasibility Study Protocol for Socrates 2.0: A Novel Cognitive Behavioral Therapy-Based Generative AI Tool to Facilitate Socratic Dialogue (Preprint) DOI Creative Commons
Philip Held, Sarah Pridgen,

Yaozhong Chen

и другие.

Опубликована: Март 8, 2024

BACKGROUND Digital mental health tools, designed to augment traditional treatments, are becoming increasingly important due a wide range of barriers accessing care, including growing shortage clinicians. Most existing tools use rule-based algorithms, often leading interactions that feel unnatural compared with human therapists. Large language models (LLMs) offer solution for the development more natural, engaging digital tools. In this paper, we detail Socrates 2.0, which was engage users in Socratic dialogue surrounding unrealistic or unhelpful beliefs, core technique cognitive behavioral therapies. The multiagent LLM-based tool features an artificial intelligence (AI) therapist, Socrates, receives automated feedback from AI supervisor and rater. combination multiple agents appeared help address common LLM issues such as looping, it improved overall experience. Initial user individuals lived experiences problems well therapists has been positive. Moreover, tests approximately 500 scenarios showed 2.0 engaged harmful responses under 1% cases, promptly correcting each time. However, formal feasibility studies potential end needed. OBJECTIVE This mixed methods study examines 2.0. METHODS On basis initial data, devised gather qualitative quantitative data about users’ clinicians’ experience interacting tool. Using method approach, goal is acceptability 100 50 clinicians inform eventual implementation generative treatment. We better understand how interact tool, frequency, length, time interactions, satisfaction overall, quality individual responses, ways should be before used efficacy trials. Descriptive inferential analyses will performed on validated usability measures. Thematic analysis data. RESULTS Recruitment begin February 2024 expected conclude by 2025. As September 25, 2024, 55 participants have recruited. CONCLUSIONS outlined first steps applying treatment delivery lay foundation studies. INTERNATIONAL REGISTERED REPORT DERR1-10.2196/58195

Язык: Английский

Процитировано

0
Formulation as Representation - Modelling the Cognitive Space of Mental Health Clinical Reasoning DOI Open Access
Andrew Hider, Su Hua Sim

Опубликована: Март 16, 2024

This paper discusses the specific characteristics of any hypothetical cognitive space that may be modelled in order to automate (or partially automate) kind mental health clinical reasoning - or psychological case formulation is used by professionals. It argues work into use generative artificial intelligence (AI) field needs consider three components this reasoning. Firstly, heterotopy. When statements are made, parsing them does not result same representation when words used, due fact ontologies contain multiple meanings for words. Secondly, orthogonality. Variables relevant causally intersect but both and treatment determination. Thirdly, veridicality. The truth a determined testable observations. Even response allow determination truth., status hinge principally on degree which it confers meaning understanding state person who experiencing state, different judgements healthcare clinician. Automated models need accommodate these features formulation.

Язык: Английский

Процитировано

0