Yuri Quintana
@yuriquintana.com
Chief Clinical Informatics @ BIDMC & Faculty Harvard Medical School, Chief of @DCINetwork.org #MedicalInformatics #AI #DigitalHealth #MobileHealth #LearningHealthSystems
https://www.LinkedIn.com/in/yuriquintana
http://www.yuriquintana.com
https://www.LinkedIn.com/in/yuriquintana
http://www.yuriquintana.com
Pinned
Call for Submissions – Patient-Powered Digital Health 2026 Conference. Papers will be in PubMed-indexed proceedings. The conference brings together patients, innovators, industry, healthcare providers, thought leaders. Due March 10, 2026. Early discount open now www.dcinetwork.org/patients2026...
apparently not just other countries should have the right to freedom of expression. The First Amendment is still in effect in this country. Don't take this for granted. Elections matter. Your voice matters. Democracy needs constitution, laws, freedom of exp. storage.courtlistener.com/recap/gov.us...
Order on Motion for TRO – #85 in Tincher v. Noem (D. Minnesota, 0:25-cv-04669) – CourtListener.com
ORDER granting in part and denying in part 16 Motion for Preliminary Injunction. (Written Opinion) Signed by Judge Katherine M. Menendez on 1/16/2026. (Entered: 01/16/2026)
storage.courtlistener.com
January 17, 2026 at 2:24 PM
apparently not just other countries should have the right to freedom of expression. The First Amendment is still in effect in this country. Don't take this for granted. Elections matter. Your voice matters. Democracy needs constitution, laws, freedom of exp. storage.courtlistener.com/recap/gov.us...
A new brain inspired memory system that learns routes like the human hippocampus. Using low power neuromorphic hardware, it recalls paths even with noise. The work shows how future AI may grow by combining many brain like modules, not one giant model. www.research-collection.ethz.ch/server/api/c...
www.research-collection.ethz.ch
January 17, 2026 at 2:10 PM
A new brain inspired memory system that learns routes like the human hippocampus. Using low power neuromorphic hardware, it recalls paths even with noise. The work shows how future AI may grow by combining many brain like modules, not one giant model. www.research-collection.ethz.ch/server/api/c...
Data contamination threatens #LLM #AIEvaluation
Scaling has “limits to growth”. New #ARCAGI2 counters this problem with contamination resistant, compositional reasoning tests and human baselines require original reasoning Not just memory recall evaluation arxiv.org/abs/2505.11831
Scaling has “limits to growth”. New #ARCAGI2 counters this problem with contamination resistant, compositional reasoning tests and human baselines require original reasoning Not just memory recall evaluation arxiv.org/abs/2505.11831
ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems
The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI), introduced in 2019, established a challenging benchmark for evaluating the general fluid intelligence of artificial ...
arxiv.org
January 17, 2026 at 1:55 PM
Data contamination threatens #LLM #AIEvaluation
Scaling has “limits to growth”. New #ARCAGI2 counters this problem with contamination resistant, compositional reasoning tests and human baselines require original reasoning Not just memory recall evaluation arxiv.org/abs/2505.11831
Scaling has “limits to growth”. New #ARCAGI2 counters this problem with contamination resistant, compositional reasoning tests and human baselines require original reasoning Not just memory recall evaluation arxiv.org/abs/2505.11831
#AIGovernance #Sustainability #SystemsThinking 2025 paper models AI scaling with “limits to growth” dynamics across technical, economic, ecological, social lenses. It argues growth is temporarily sustained by shifting constraints, often externalizing harms. arxiv.org/abs/2501.17980
Limits to AI Growth: The Ecological and Social Consequences of Scaling
The accelerating development and deployment of AI technologies depend on the continued ability to scale their infrastructure. This has implied increasing amounts of monetary investment and natural res...
arxiv.org
January 17, 2026 at 1:34 PM
#AIGovernance #Sustainability #SystemsThinking 2025 paper models AI scaling with “limits to growth” dynamics across technical, economic, ecological, social lenses. It argues growth is temporarily sustained by shifting constraints, often externalizing harms. arxiv.org/abs/2501.17980
#DataContamination #AIEvaluation Training–test overlap can inflate LLM scores. “data contamination” in #LLMs, defined as unintended overlap between training data & evaluation data that can inflate measured performance & misrepresent true generalization. arxiv.org/html/2502.14...
A Survey on Data Contamination for Large Language Models
arxiv.org
January 17, 2026 at 1:32 PM
#DataContamination #AIEvaluation Training–test overlap can inflate LLM scores. “data contamination” in #LLMs, defined as unintended overlap between training data & evaluation data that can inflate measured performance & misrepresent true generalization. arxiv.org/html/2502.14...
#PancreaticCancer #AI #PrecisionOncology Researchers used the catRAPID AI platform to design Apt1, a DNA aptamer that blocks RAD51–BRCA2 repair signaling, disabling homologous recombination and restoring sensitivity to olaparib in pancreatic cancer cells and 3D models. doi.org/10.1038/s414...
Client Challenge
doi.org
January 15, 2026 at 3:13 AM
#PancreaticCancer #AI #PrecisionOncology Researchers used the catRAPID AI platform to design Apt1, a DNA aptamer that blocks RAD51–BRCA2 repair signaling, disabling homologous recombination and restoring sensitivity to olaparib in pancreatic cancer cells and 3D models. doi.org/10.1038/s414...
Interview: The arrival of #AGI | Shane Legg (co-founder of DeepMind) #MLSKY - See www.youtube.com/watch?v=l3u_...
The arrival of AGI | Shane Legg (co-founder of DeepMind)
YouTube video by Google DeepMind
www.youtube.com
January 13, 2026 at 3:47 PM
Interview: The arrival of #AGI | Shane Legg (co-founder of DeepMind) #MLSKY - See www.youtube.com/watch?v=l3u_...
OpenAI bought Torch to improve ChatGPT Health. It connects medical records plus wellness apps such as Apple Health & Peloton, lets users ask questions about their data while keeping it encrypted & separate from regular chats. OpenAI's pushes into healthcare. torchapp.com
Torch - Track Medical Records
Torch makes sense of your scattered health data. Connect thousands of sources and get clear, beautiful answers to your most important health questions with AI.
torchapp.com
January 13, 2026 at 3:42 PM
OpenAI bought Torch to improve ChatGPT Health. It connects medical records plus wellness apps such as Apple Health & Peloton, lets users ask questions about their data while keeping it encrypted & separate from regular chats. OpenAI's pushes into healthcare. torchapp.com
Call for Submissions – Patient-Powered Digital Health 2026 Conference. Papers will be in PubMed-indexed proceedings. The conference brings together patients, innovators, industry, healthcare providers, thought leaders. Due March 10, 2026. Early discount open now www.dcinetwork.org/patients2026...
January 13, 2026 at 12:10 AM
Call for Submissions – Patient-Powered Digital Health 2026 Conference. Papers will be in PubMed-indexed proceedings. The conference brings together patients, innovators, industry, healthcare providers, thought leaders. Due March 10, 2026. Early discount open now www.dcinetwork.org/patients2026...
#MLSky #ClinicalTrials #AI #Anthropic has launched a ClinicalTrials.gov connector for Claude, giving AI structured access to 500,000+ registered trials from the NIH, including endpoints, eligibility criteria, sponsors, phases, and recruitment status.
docs.mcp.deepsense.ai/guides/clini...
docs.mcp.deepsense.ai/guides/clini...
Clinical Trials Connector
The Clinical Trials Connector gives Claude access to ClinicalTrials.gov, the NIH/NLM registry of FDA-regulated clinical studies conducted worldwide. What You Can Do: With this connector, Claude can...
docs.mcp.deepsense.ai
January 12, 2026 at 8:44 PM
#MLSky #ClinicalTrials #AI #Anthropic has launched a ClinicalTrials.gov connector for Claude, giving AI structured access to 500,000+ registered trials from the NIH, including endpoints, eligibility criteria, sponsors, phases, and recruitment status.
docs.mcp.deepsense.ai/guides/clini...
docs.mcp.deepsense.ai/guides/clini...
Weill Cornell" whole-genome sequencing to identify more patients who can benefit from PARP inhibitor drugs. New algorithm found that nearly 25% of patients without traditional genetic mutations (like BRCA) could still benefit from these targeted treatments news.weill.cornell.edu/news/2026/01...
Whole-Genome Sequencing May Optimize PARP Inhibitor Use
A whole-genome sequencing approach shows early promise over current commercial methods for identifying more patients likely to benefit from PARP inhibitor cancer treatments, according to a study led b...
news.weill.cornell.edu
January 12, 2026 at 1:26 PM
Weill Cornell" whole-genome sequencing to identify more patients who can benefit from PARP inhibitor drugs. New algorithm found that nearly 25% of patients without traditional genetic mutations (like BRCA) could still benefit from these targeted treatments news.weill.cornell.edu/news/2026/01...
#AI #LongContext #MITCSAIL #mlsky
MIT CSAIL proposes Recursive Language Models that treat long prompts as an external environment. The model writes Python3 to inspect, decompose, and recursively query itself, enabling 10M+ token inputs. On BrowseComp+ (1K) it hits 91.33%
arxiv.org/abs/2512.24601
MIT CSAIL proposes Recursive Language Models that treat long prompts as an external environment. The model writes Python3 to inspect, decompose, and recursively query itself, enabling 10M+ token inputs. On BrowseComp+ (1K) it hits 91.33%
arxiv.org/abs/2512.24601
Recursive Language Models
We study allowing large language models (LLMs) to process arbitrarily long prompts through the lens of inference-time scaling. We propose Recursive Language Models (RLMs), a general inference strategy...
arxiv.org
January 12, 2026 at 1:02 PM
#AI #LongContext #MITCSAIL #mlsky
MIT CSAIL proposes Recursive Language Models that treat long prompts as an external environment. The model writes Python3 to inspect, decompose, and recursively query itself, enabling 10M+ token inputs. On BrowseComp+ (1K) it hits 91.33%
arxiv.org/abs/2512.24601
MIT CSAIL proposes Recursive Language Models that treat long prompts as an external environment. The model writes Python3 to inspect, decompose, and recursively query itself, enabling 10M+ token inputs. On BrowseComp+ (1K) it hits 91.33%
arxiv.org/abs/2512.24601
Anthropic joins OpenAI's push into health care with new Claude tools Anthropic’s new offerings allow users and providers to work with medical data, mimicking similar moves by OpenAI www.nbcnews.com/tech/tech-ne...
Anthropic joins OpenAI's push into health care with new Claude tools
Anthropic’s new offerings allow users and providers to work with medical data, mimicking similar moves by OpenAI.
www.nbcnews.com
January 12, 2026 at 4:36 AM
Anthropic joins OpenAI's push into health care with new Claude tools Anthropic’s new offerings allow users and providers to work with medical data, mimicking similar moves by OpenAI www.nbcnews.com/tech/tech-ne...
Editorial from 50 former NIH staff - "The NIH has lost its scientific integrity. So we left - We tried to be optimistic about the Trump administration. But we could not continue" www.statnews.com/2026/01/10/n...
The NIH has lost its scientific integrity. So we left
“We can no longer lend our credibility to an organization that has lost its integrity,” write four scientists and administrators who recently resigned from the NIH.
www.statnews.com
January 11, 2026 at 4:41 PM
Editorial from 50 former NIH staff - "The NIH has lost its scientific integrity. So we left - We tried to be optimistic about the Trump administration. But we could not continue" www.statnews.com/2026/01/10/n...
#AI #Safety #LLMDefense #Anthropic introduces Constitutional Classifiers++: exchange-aware filtering plus a low-cost activation probe cascade that escalates only suspicious chats, reducing universal jailbreak success while keeping refusals rare, with ~1% compute overhead arxiv.org/pdf/2601.046...
arxiv.org
January 11, 2026 at 12:37 AM
#AI #Safety #LLMDefense #Anthropic introduces Constitutional Classifiers++: exchange-aware filtering plus a low-cost activation probe cascade that escalates only suspicious chats, reducing universal jailbreak success while keeping refusals rare, with ~1% compute overhead arxiv.org/pdf/2601.046...
#AIinHealthcare #PrescriptionAI #HealthPolicy Utah launches first US pilot for AI chatbot prescription renewals for 200 common drugs, easing doctor workloads and improving access. www.washingtonpost.com/nation/2026/...
Utah launches first-in-the-nation trial that lets AI renew your prescription
Utah said that with its new AI prescription drug chatbot, it wants to free up doctors to focus on patient care and make medicine more accessible and affordable.
www.washingtonpost.com
January 10, 2026 at 12:06 PM
#AIinHealthcare #PrescriptionAI #HealthPolicy Utah launches first US pilot for AI chatbot prescription renewals for 200 common drugs, easing doctor workloads and improving access. www.washingtonpost.com/nation/2026/...
When millions turn to AI for health guidance, transparency isn't optional; it's essential. We are at a crossroads. The choices we make now will determine if AI empowers patients through transparent, accountable systems or leaves them navigating invisible risks. www.linkedin.com/pulse/chatgp...
ChatGPT Health Launches—But Healthcare AI Needs More Than Promises
OpenAI's announcement of ChatGPT Health (https://openai.com/index/introducing-chatgpt-health/) marks a significant moment in healthcare AI.
www.linkedin.com
January 8, 2026 at 8:24 PM
When millions turn to AI for health guidance, transparency isn't optional; it's essential. We are at a crossroads. The choices we make now will determine if AI empowers patients through transparent, accountable systems or leaves them navigating invisible risks. www.linkedin.com/pulse/chatgp...
ChatGPT Health — a dedicated space for health conversations in ChatGPT. You can securely connect medical records and wellness apps so responses are grounded in your own health information. openai.com/index/introd...
Introducing ChatGPT Health
A dedicated experience in ChatGPT designed for health and wellness.
openai.com
January 8, 2026 at 3:38 AM
ChatGPT Health — a dedicated space for health conversations in ChatGPT. You can securely connect medical records and wellness apps so responses are grounded in your own health information. openai.com/index/introd...
#AI #Privacy #Switzerland Euria is a sovereign Swiss AI assistant hosted in Switzerland, powered by renewable energy, with no user data used for training and an ephemeral mode for zero retention. Server heat is recycled to warm homes in Geneva. euria.infomaniak.com 
Euria by Infomaniak
Meet Euria, the sovereign AI that protects your privacy and uses its energy to heat homes.
euria.infomaniak.com
January 8, 2026 at 1:12 AM
#AI #Privacy #Switzerland Euria is a sovereign Swiss AI assistant hosted in Switzerland, powered by renewable energy, with no user data used for training and an ephemeral mode for zero retention. Server heat is recycled to warm homes in Geneva. euria.infomaniak.com 
#HealthcareAI #DigitalHealth #OpenAI Jan 2026: over 5% of ChatGPT messages are healthcare related and 25% of weekly active users ask health questions, emphasizing navigation and safety. cdn.openai.com/pdf/2cb29276...
Source: OpenAI, AI as a Healthcare Ally (January 2026). 
Source: OpenAI, AI as a Healthcare Ally (January 2026). 
cdn.openai.com
January 6, 2026 at 8:31 PM
#HealthcareAI #DigitalHealth #OpenAI Jan 2026: over 5% of ChatGPT messages are healthcare related and 25% of weekly active users ask health questions, emphasizing navigation and safety. cdn.openai.com/pdf/2cb29276...
Source: OpenAI, AI as a Healthcare Ally (January 2026). 
Source: OpenAI, AI as a Healthcare Ally (January 2026). 
#MedicalAI #AITrust #UncertaintyMap BMJ Paper introduces U-map linking AI uncertainties to clinical roles, proposing posthuman collaboration for trust in medical applications.
mh.bmj.com/content/earl...
mh.bmj.com/content/earl...
Spectres of medical AI: uncertainty, trust and the posthuman condition
Artificial intelligence (AI) is transforming clinical practice while simultaneously raising concerns about trust. Drawing on complexity theory, this paper argues that the crisis of trust in medical AI...
mh.bmj.com
January 6, 2026 at 2:14 PM
#MedicalAI #AITrust #UncertaintyMap BMJ Paper introduces U-map linking AI uncertainties to clinical roles, proposing posthuman collaboration for trust in medical applications.
mh.bmj.com/content/earl...
mh.bmj.com/content/earl...
#AIinMentalHealth #AIEmpathy #EthicalAI Uganda chatbot capable of detecting mental health issues in local languages , Swahili and Luganda, shows LLMs outperform humans in empathy tasks, warns of redefining empathy, risks in psychosis in vulnerable users with #mhapps
www.ft.com/content/ef77...
www.ft.com/content/ef77...
The problem with AI and ‘empathy’
If technology redefines what our language means it could also change our perceptions of ourselves
www.ft.com
January 6, 2026 at 2:13 PM
#AIinMentalHealth #AIEmpathy #EthicalAI Uganda chatbot capable of detecting mental health issues in local languages , Swahili and Luganda, shows LLMs outperform humans in empathy tasks, warns of redefining empathy, risks in psychosis in vulnerable users with #mhapps
www.ft.com/content/ef77...
www.ft.com/content/ef77...
#AIinCriticalCare #PediatricICU #HealthcareAI JAMA Viewpoint advocates AI for de-escalating care in pediatric ICUs, leveraging physiologic data for predictions amid limited clinical adoption despite model proliferation. jamanetwork.com/journals/jam...
AI in Critical Care—Use for De-Escalation Rather Than Escalation
This Viewpoint examines the gap between the development of artificial intelligence (AI) models in pediatric intensive care and their use in clinical practice and proposes deployment of AI models durin...
jamanetwork.com
January 6, 2026 at 2:10 PM
#AIinCriticalCare #PediatricICU #HealthcareAI JAMA Viewpoint advocates AI for de-escalating care in pediatric ICUs, leveraging physiologic data for predictions amid limited clinical adoption despite model proliferation. jamanetwork.com/journals/jam...
#BehavioralHealth #AIGovernance Nature perspective calls for empowering service users, public, and providers in AI decision-making for behavioral healthcare, countering private sector dominance amid billions in investments and ethical concerns like job displacement.
www.nature.com/articles/s44...
www.nature.com/articles/s44...
Empowering service users, the public, and providers to determine the future of artificial intelligence in behavioral healthcare - Nature Mental Health
In this Perspective, Last and Khazanov call for democratizing AI in behavioral healthcare, urging that service users, providers and the public—not private interests—shape its development and deploymen...
www.nature.com
January 6, 2026 at 2:08 PM
#BehavioralHealth #AIGovernance Nature perspective calls for empowering service users, public, and providers in AI decision-making for behavioral healthcare, countering private sector dominance amid billions in investments and ethical concerns like job displacement.
www.nature.com/articles/s44...
www.nature.com/articles/s44...
#MLSKY Claude Scientific Skills open source repository from K Dense, library of “skills” turn Claude into practical scientific research assistant, multistep workflows, life sciences & clinical research, standardizes how Claude connects to scientific software & data sources github.com/K-Dense-AI/c...
GitHub - K-Dense-AI/claude-scientific-skills: A set of ready to use scientific skills for Claude
A set of ready to use scientific skills for Claude - K-Dense-AI/claude-scientific-skills
github.com
January 4, 2026 at 4:34 PM
#MLSKY Claude Scientific Skills open source repository from K Dense, library of “skills” turn Claude into practical scientific research assistant, multistep workflows, life sciences & clinical research, standardizes how Claude connects to scientific software & data sources github.com/K-Dense-AI/c...