iiinova — where your digital self keeps evolving

Emre Sezgin, PhD @emreszg.bsky.social · 26/10/2025

Attention is all you need, but misdirected. Sharing my new curmudgeon corner piece at AI & Society reflecting my thoughts on emerging societal issue on our attentions rdcu.be/eMLDy

rdcu.be

Attention is all you need? When responsiveness short-circuits responsibility

000 0

Reposted by Emre Sezgin, PhD

Alvin @alviebarr.bsky.social · 01/07/2025

EXCITED TO SHARE! Here we show synthetic clinical data may be used in place of real-world data to answer research questions, train ML models, and to amplify sample sizes Check it out: thejns.org/focus/view/j... @emreszg.bsky.social

064 0

Emre Sezgin, PhD @emreszg.bsky.social · 08/04/2025

Tomorrow at CCHEOR C.A.R.E.S, I’ll present new results from our study on an AI chatbot helping low-income families to navigate social needs land connect with community resources. Register: nationwidechildrens.cloud-cme.com/course/cours... @nationwidekids.bsky.social #DigitalHealth #conversationalAI

010 0

Emre Sezgin, PhD @emreszg.bsky.social · 29/03/2025

Wrapping up an inspiring week at #SBM2025, connecting w/ researchers driving innovation in behavioral science & digital health. Also, I was honored to receive the SBM Digital Health SIG Award for Excellence in Digital Health Publication—a humbling recognition + motivation to keep up the momentum!

040 0

Emre Sezgin, PhD @emreszg.bsky.social · 28/03/2025

It is poster time! Find me at D77 to learn more about our chatbot for helping caregivers of young childhood survivors at #SBM2025 @sbm.org

050 0

Emre Sezgin, PhD @emreszg.bsky.social · 26/03/2025

On my way to #SBM2025 I’ll be presenting our work on using conversational AI to support caregivers of childhood cancer survivors.

000 0

Reposted by Emre Sezgin, PhD

Ethan Mollick @emollick.bsky.social · 21/03/2025

New controlled study of AI voice mode on mental health finds complex results: GPT-4 with an empathetic voice reduced loneliness, but heavier users, and especially when use was combined with a more neutral voice or those with particular personalities, had multiple negative impacts on users.

34812 0

Emre Sezgin, PhD @emreszg.bsky.social · 24/02/2025

I have a knack for understanding if a human or a machine is "particularly impressed by" my recent article... #academicLife

000 0

Reposted by Emre Sezgin, PhD

Alvin @alviebarr.bsky.social · 14/02/2025

Super excited to share our latest preprint: We generate synthetic neurosurgical data with an LLM (using a single prompt) and train an ML model on it to predict real-world neurosurgical outcomes Read more: arxiv.org/pdf/2502.09566 @emreszg.bsky.social

0124 0

Reposted by Emre Sezgin, PhD

Alvin @alviebarr.bsky.social · 06/02/2025

Excited to share the first in our series of papers showcasing the ability to generate synthetic clinical datasets with LLMs, using zero-shot prompting Read more: www.frontiersin.org/journals/art... @emreszg.bsky.social

043 0

Emre Sezgin, PhD @emreszg.bsky.social · 07/01/2025

🔗Bing Chat and Google SGE frequently linked to recent, expert-created resources, enhancing their credibility. 💬 Bard consistently included AI disclosures, promoting transparency, while ChatGPT did so less frequently. 📌 ChatGPT showed high content originality.

000 0

Emre Sezgin, PhD @emreszg.bsky.social · 07/01/2025

All models scored higher than the recommended 8th grade readability level for health information, with ChatGPT and Bard providing more complex responses. Google SGE often provided shorter, more concise answers

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 07/01/2025

ChatGPT performed relatively better than others in accuracy, clarity, completeness, clinical utility, and overall rating. Bing Chat and Google SGE had lower overall scores, highlighting the areas for improvement among the models.

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 07/01/2025

🔍 We evaluated how well LLMs respond to caregivers of children with cancer in finding reliable information, focusing on 4 LLM-based services available: ChatGPT (GPT-4), Google Bard (Gemini Pro), Microsoft Bing Chat, and Google SGE (Google's AI integration to search engine).

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 19/12/2024

With chatGPT on WhatsApp and phone calls, as well as meta AI, My colleague @abakik.bsky.social and I have been discussing the huge potential of such integration of generalist conversational AI to current messaging services in public health. More details are on our preprint @jmirpub.bsky.social

preprints.jmir.org

The Era of Generalist Conversational AI to Support Public Health Communications

The integration of artificial intelligence (AI) into health communication systems has introduced a transformative approach to public health management, particularly during health emergencies....

021 0

Reposted by Emre Sezgin, PhD

Eric Topol @erictopol.bsky.social · 18/12/2024

Addressing #AI bias and lack of transparency for healthcare datasets Consensus recommendations from a global consortium thelancet.com/journals/lan...

36723 0

Emre Sezgin, PhD @emreszg.bsky.social · 19/12/2024

As AI generated content has been used all around our digital ecosystem, I started to appreciate typos a bit more, an indication of human touch, until someone prompt chatgpt to add typos...

010 0

Emre Sezgin, PhD @emreszg.bsky.social · 11/12/2024

This isn’t just about tech—it’s about people. Let’s rethink how AI can truly support healthcare professionals.

000 0

Emre Sezgin, PhD @emreszg.bsky.social · 11/12/2024

Key considerations: 💻 Are AI tools functional, intuitive, and well-integrated? 👂 Are we listening to radiologists about their needs, or prioritizing high-accuracy tasks without regard to usability? 🌐 Is organizational culture ready for AI adoption?

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 11/12/2024

This sparks crucial questions about how AI tools are designed, integrated, and used in real-world clinical practice. Are we truly helping providers, or adding complexity to their work? 🩺🤖

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

While there are scenarios where AI can play an essential role—such as in low-resource environments with limited provider availability—it’s crucial to emphasize these realities without exacerbating fear and anxiety among already overburdened professionals.

000 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

One of the earlier NYT articles on "AI to replace radiologists" last year nudged me to write a commentary last year (with a similar titling;) ), advocating for collaboration between AI and providers rather than framing the narrative as one "beating" the other: journals.sagepub.com/doi/10.1177/...

journals.sagepub.com

Sage Journals: Discover world-class research

Subscription and open access journals from Sage, the world's leading independent academic publisher.

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

This kind of sensationalism can unnecessarily heighten anxiety among healthcare providers about job security, adding to the significant stress they already face (as evidenced by research and even the community discussions on platforms like Reddit posts).

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

This can lead to misconceptions. Studies often rely on controlled assumptions and specific criteria. However, these scenarios rarely reflect real-world conditions, which require organizational decisions, strategies, navigating health systems &eventually human involvement for safe & effective use.

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

It is always a bigger risk with titles than article content as many readers focus only on these headlines to deduct to the conclusion, with fewer reading the full article, and an even smaller subset examining the underlying research papers or their assumptions and limitations.

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 21/11/2024

The New York Times has once again chosen a provocative headline, a common but risky strategy they use to capture attention. While this approach may drive clicks, it risks fostering an imbalanced trust in AI at the expense of healthcare providers.

100 0

Emre Sezgin, PhD @emreszg.bsky.social · 18/11/2024

I am happy to share our first publication from DAPHNE project! Our study highlights the design and evaluation of our DAPHNE chatbot, aimed at screening and understanding social needs and linking patients and families to essential resources. humanfactors.jmir.org/2024/1/e57114/ #digitalhealth

humanfactors.jmir.org

Chatbot for Social Need Screening and Resource Sharing With Vulnerable Families: Iterative Design and Evaluation Study

Background: Health outcomes are significantly influenced by unmet social needs. Although screening for social needs has become common in health care settings, there is often poor linkage to resources ...

010 0