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
AlphaEvolve integrates large language models with evolutionary algorithms to autonomously generate, test, and refine mathematical constructions. It rediscovered known results, improved others, and points to a future of AI-human collaboration in mathematical discovery. arxiv.org/abs/2511.02864
Mathematical exploration and discovery at scale
AlphaEvolve is a generic evolutionary coding agent that combines the generative capabilities of LLMs with automated evaluation in an iterative evolutionary framework that proposes, tests, and refines ...
arxiv.org
November 9, 2025 at 11:50 PM
AlphaEvolve integrates large language models with evolutionary algorithms to autonomously generate, test, and refine mathematical constructions. It rediscovered known results, improved others, and points to a future of AI-human collaboration in mathematical discovery. arxiv.org/abs/2511.02864
New Trump policy changes how NIH funds grants. By committing multi-year money upfront, fewer new awards will be available, making competition tougher for researchers. Read the full Forbes analysis: www.forbes.com/sites/brucel...
How New Trump Multiyear Policy Makes NIH Grants Even Harder To Get
The new Trump administration multiyear funding policy for NIH grants has dropped National Cancer Institute grant funding rates from one in 10 applicants to one in 25.
www.forbes.com
November 9, 2025 at 11:39 PM
New Trump policy changes how NIH funds grants. By committing multi-year money upfront, fewer new awards will be available, making competition tougher for researchers. Read the full Forbes analysis: www.forbes.com/sites/brucel...
McKinsey’s 2025 AI Report:
The State of AI: How Organizations Are Rewiring to Capture Value (2025)
78% use AI, but few see big gains. CEO-led AI, workflow redesign, and risk management drive real impact. Scaling still the missing link.
www.mckinsey.com/~/media/mcki...
The State of AI: How Organizations Are Rewiring to Capture Value (2025)
78% use AI, but few see big gains. CEO-led AI, workflow redesign, and risk management drive real impact. Scaling still the missing link.
www.mckinsey.com/~/media/mcki...
www.mckinsey.com
November 9, 2025 at 11:18 PM
McKinsey’s 2025 AI Report:
The State of AI: How Organizations Are Rewiring to Capture Value (2025)
78% use AI, but few see big gains. CEO-led AI, workflow redesign, and risk management drive real impact. Scaling still the missing link.
www.mckinsey.com/~/media/mcki...
The State of AI: How Organizations Are Rewiring to Capture Value (2025)
78% use AI, but few see big gains. CEO-led AI, workflow redesign, and risk management drive real impact. Scaling still the missing link.
www.mckinsey.com/~/media/mcki...
FutureHouse has launched a new generation of AI agents for science—Crow, Falcon, Owl, and Phoenix—designed to reason across papers, verify findings, and accelerate discovery with transparency and rigor. Learn more: www.futurehouse.org
FutureHouse
FutureHouse is a philanthropically-funded moonshot focused on building an AI Scientist. Our 10-year mission is to build semi-autonomous AIs for scientific research, to accelerate the pace of discovery...
www.futurehouse.org
November 9, 2025 at 7:21 PM
FutureHouse has launched a new generation of AI agents for science—Crow, Falcon, Owl, and Phoenix—designed to reason across papers, verify findings, and accelerate discovery with transparency and rigor. Learn more: www.futurehouse.org
Google just published DS-STAR, a state-of-the-art data science agent that automates everything from data wrangling to visualization. It reads messy files, plans steps, writes and verifies code, and sets new SOTA on DABStep, KramaBench, and DA-Code.
🔗 research.google/blog/ds-star...
🔗 research.google/blog/ds-star...
DS-STAR: A state-of-the-art versatile data science agent
research.google
November 9, 2025 at 3:42 AM
Google just published DS-STAR, a state-of-the-art data science agent that automates everything from data wrangling to visualization. It reads messy files, plans steps, writes and verifies code, and sets new SOTA on DABStep, KramaBench, and DA-Code.
🔗 research.google/blog/ds-star...
🔗 research.google/blog/ds-star...
HALO introduces a hierarchical framework for multi-agent LLM systems, combining planning, orchestration, and verification to improve reasoning and efficiency in complex tasks. Read the 2025 paper: arxiv.org/abs/2505.13516
HALO: Hierarchical Autonomous Logic-Oriented Orchestration for Multi-Agent LLM Systems
Recent advancements in Multi-Agent Systems (MAS) powered by Large Language Models (LLMs) have demonstrated tremendous potential in diverse task scenarios. Nonetheless, existing agentic systems typical...
arxiv.org
November 9, 2025 at 3:22 AM
HALO introduces a hierarchical framework for multi-agent LLM systems, combining planning, orchestration, and verification to improve reasoning and efficiency in complex tasks. Read the 2025 paper: arxiv.org/abs/2505.13516
HALO introduces a hierarchical framework for multi-agent LLM systems, combining planning, orchestration, and verification to improve reasoning and efficiency in complex tasks. Read the 2025 paper: arxiv.org/abs/2505.13516
HALO: Hierarchical Autonomous Logic-Oriented Orchestration for Multi-Agent LLM Systems
Recent advancements in Multi-Agent Systems (MAS) powered by Large Language Models (LLMs) have demonstrated tremendous potential in diverse task scenarios. Nonetheless, existing agentic systems typical...
arxiv.org
November 9, 2025 at 3:22 AM
HALO introduces a hierarchical framework for multi-agent LLM systems, combining planning, orchestration, and verification to improve reasoning and efficiency in complex tasks. Read the 2025 paper: arxiv.org/abs/2505.13516
SYSTEM PROMPT: REALITY FILTER
**Purpose:**
Reduce hallucinations by enforcing mechanical truth discipline — make ChatGPT explicitly identify unverifiable or inferred content rather than presenting it as fact.
**Purpose:**
Reduce hallucinations by enforcing mechanical truth discipline — make ChatGPT explicitly identify unverifiable or inferred content rather than presenting it as fact.
November 9, 2025 at 3:00 AM
SYSTEM PROMPT: REALITY FILTER
**Purpose:**
Reduce hallucinations by enforcing mechanical truth discipline — make ChatGPT explicitly identify unverifiable or inferred content rather than presenting it as fact.
**Purpose:**
Reduce hallucinations by enforcing mechanical truth discipline — make ChatGPT explicitly identify unverifiable or inferred content rather than presenting it as fact.
China’s new “Socrates AI” builds a scientific encyclopedia that explains not just what the universe is, but why. It maps 3M reasoning chains across 200 fields to reveal the logic behind science. Inverse knowledge search arrives: arxiv.org/abs/2510.26854
Inverse Knowledge Search over Verifiable Reasoning: Synthesizing a Scientific Encyclopedia from a Long Chains-of-Thought Knowledge Base
Most scientific materials compress reasoning, presenting conclusions while omitting the derivational chains that justify them. This compression hinders verification by lacking explicit, step-wise just...
arxiv.org
November 9, 2025 at 2:45 AM
China’s new “Socrates AI” builds a scientific encyclopedia that explains not just what the universe is, but why. It maps 3M reasoning chains across 200 fields to reveal the logic behind science. Inverse knowledge search arrives: arxiv.org/abs/2510.26854
Reposted by Yuri Quintana
November 7, 2025 at 6:13 PM
Or be more like Canadians and a Japanese society who citizens have less guns and have better legislation. Sadly in the USA the Dickey Amendment says “none of the funds made available for injury prevention and control at the CDC may be used to advocate or promote gun control.” Science is muzzled
Two more families lives ruined by idiots with guns. One of the four cardinal rules of firearm safety is, “know your target and what is behind it.“. I wonder if we would actually be a safer country if everyone learned firearm safety in public schools. www.nytimes.com/2025/11/06/u...
Cleaning Woman Killed Through Door After Arriving at Wrong Home, Police Say
www.nytimes.com
November 8, 2025 at 2:51 PM
Or be more like Canadians and a Japanese society who citizens have less guns and have better legislation. Sadly in the USA the Dickey Amendment says “none of the funds made available for injury prevention and control at the CDC may be used to advocate or promote gun control.” Science is muzzled
Reposted by Yuri Quintana
The use of GenAI in health research necessitates transparent reporting standards.
New guidelines help: CHART for chatbot health advice, TRIPOD-LLM for model development or prediction, and GAMER for GenAI-assisted manuscript writing.
#MedSky #MedAI #MLSky
New guidelines help: CHART for chatbot health advice, TRIPOD-LLM for model development or prediction, and GAMER for GenAI-assisted manuscript writing.
#MedSky #MedAI #MLSky
Reporting guidelines for studies involving generative artificial intelligence applications: what do I use, and when? - npj Digital Medicine
With a growing number of studies applying generative artificial intelligence (GAI) models for health purposes, reporting standards are being developed to guide authors in this space. We describe the…
www.nature.com
November 7, 2025 at 4:09 PM
Reposted by Yuri Quintana
🧪 Epic Research leverages its vast Cosmos database for rapid, direct-to-web health studies, skipping traditional peer review. This model trades speed for external vetting, raising concerns about rigor even as its findings influence policy.
#MedSky #Research
#MedSky #Research
Inside Epic’s Rapid-Fire Research Machine
Epic Research is rewriting the rules of medical publishing by using massive patient datasets to deliver fast, accessible studies outside traditional journals.
www.medscape.com
November 7, 2025 at 5:07 PM
Reposted by Yuri Quintana
share.google/Cx9LQMsTaXbO...
This is an interesting way to look at this.
This is an interesting way to look at this.
Fitbit's latest experiment lets Pixel Watch users check for early signs of bad health
Fitbit Labs now allows users try out upcoming Pixel Watch features such as Unusual Health Detection and hypertension alerts.
share.google
November 7, 2025 at 11:51 PM
share.google/Cx9LQMsTaXbO...
This is an interesting way to look at this.
This is an interesting way to look at this.
I am deeply honored to be elected as a Fellow of the American Medical Informatics Association (FAMIA). This reflects collective achievements of my research team & collaborators, whose dedication to advancing healthcare through informatics continues to inspire & motivate me. amia.org/news-publica...
Congratulations 2026 FAMIA Inductees
Congratulations to the 2026 class of Fellows of
amia.org
November 6, 2025 at 7:11 PM
I am deeply honored to be elected as a Fellow of the American Medical Informatics Association (FAMIA). This reflects collective achievements of my research team & collaborators, whose dedication to advancing healthcare through informatics continues to inspire & motivate me. amia.org/news-publica...
Google's rolling out its most powerful AI chip, taking aim at Nvidia with custom silicon www.cnbc.com/2025/11/06/g...
Google's rolling out its most powerful AI chip, taking aim at Nvidia with custom silicon
Google says Ironwood, the seventh generation of its Tensor Processing Unit, is more than four times faster than its predecessor chip.
www.cnbc.com
November 6, 2025 at 7:09 PM
Google's rolling out its most powerful AI chip, taking aim at Nvidia with custom silicon www.cnbc.com/2025/11/06/g...
TransferEngine delivers portable RDMA point to point communication for large language model systems achieving 400 Gbps on ConnectX 7 and AWS EFA while enabling trillion parameter updates in 1.3 seconds. arxiv.org/abs/2510.27656
RDMA Point-to-Point Communication for LLM Systems
Emerging Large Language Model (LLM) system patterns, such as disaggregated inference, Mixture-of-Experts (MoE) routing, and asynchronous reinforcement fine-tuning, require flexible point-to-point comm...
arxiv.org
November 6, 2025 at 5:51 AM
TransferEngine delivers portable RDMA point to point communication for large language model systems achieving 400 Gbps on ConnectX 7 and AWS EFA while enabling trillion parameter updates in 1.3 seconds. arxiv.org/abs/2510.27656
Kosmos introduces a new era of AI-driven science by autonomously performing literature review, hypothesis generation, and data analysis. It produced reproducible, novel findings across multiple domains. Read the full paper arxiv.org/abs/2511.02824
Kosmos: An AI Scientist for Autonomous Discovery
Data-driven scientific discovery requires iterative cycles of literature search, hypothesis generation, and data analysis. Substantial progress has been made towards AI agents that can automate scient...
arxiv.org
November 6, 2025 at 5:39 AM
Kosmos introduces a new era of AI-driven science by autonomously performing literature review, hypothesis generation, and data analysis. It produced reproducible, novel findings across multiple domains. Read the full paper arxiv.org/abs/2511.02824
Data scientists perform last rites for 'dearly departed datasets' in 2nd Trump administration
apnews.com/article/cens...
apnews.com/article/cens...
Data scientists perform last rites for 'dearly departed datasets' in 2nd Trump administration
A group of U.S. data scientists has published a list of federal datasets that have been altered or removed since President Donald Trump returned to the White House.
apnews.com
November 6, 2025 at 5:30 AM
Data scientists perform last rites for 'dearly departed datasets' in 2nd Trump administration
apnews.com/article/cens...
apnews.com/article/cens...
Claude for Life Sciences is a specialized AI offering from Anthropic, launched on October 20, 2025, designed to accelerate scientific discovery & research workflows in the biotech, pharmaceutical, & healthcare sectors www.anthropic.com/news/claude-...
Hypothesis Generation, Lit Review, Data Analysis
Hypothesis Generation, Lit Review, Data Analysis
Synapse | Sage Bionetworks
Synapse is a platform for supporting scientific collaborations centered around shared biomedical data sets. Our goal is to make biomedical research more transparent, more reproducible, and more ac...
Synapse.org
November 4, 2025 at 10:25 AM
Claude for Life Sciences is a specialized AI offering from Anthropic, launched on October 20, 2025, designed to accelerate scientific discovery & research workflows in the biotech, pharmaceutical, & healthcare sectors www.anthropic.com/news/claude-...
Hypothesis Generation, Lit Review, Data Analysis
Hypothesis Generation, Lit Review, Data Analysis
A great new book into what is happening to science now. In hardcover, ebook or audio book. “Science Under Siege
How to Fight the Five Most Powerful Forces that Threaten Our World”
By Michael E. Mann
and Peter J. Hotez www.hachettebookgroup.com/titles/micha...
How to Fight the Five Most Powerful Forces that Threaten Our World”
By Michael E. Mann
and Peter J. Hotez www.hachettebookgroup.com/titles/micha...
November 4, 2025 at 10:00 AM
A great new book into what is happening to science now. In hardcover, ebook or audio book. “Science Under Siege
How to Fight the Five Most Powerful Forces that Threaten Our World”
By Michael E. Mann
and Peter J. Hotez www.hachettebookgroup.com/titles/micha...
How to Fight the Five Most Powerful Forces that Threaten Our World”
By Michael E. Mann
and Peter J. Hotez www.hachettebookgroup.com/titles/micha...
Google explores how to move beyond the traditional request response model with a real time bidirectional streaming architecture for multi agent systems enabling persistent sessions and low latency interaction developers.googleblog.com/en/beyond-re...
Beyond Request-Response: Architecting Real-time Bidirectional Streaming Multi-agent System- Google Developers Blog
Architecting next-gen multi-agent AI: Ditch request-response for real-time bidi-streaming. ADK provides the architecture for concurrency and seamless agent handoffs.
developers.googleblog.com
November 4, 2025 at 9:41 AM
Google explores how to move beyond the traditional request response model with a real time bidirectional streaming architecture for multi agent systems enabling persistent sessions and low latency interaction developers.googleblog.com/en/beyond-re...
“Effective context engineering for AI agents” by Anthropic down best practices (Compaction, agentic memory, multi-agent) architectures) for composing high-quality context into components for LLM prompts for “just-in-time” loading rather than bulk everything upfront www.anthropic.com/engineering/...
Effective context engineering for AI agents
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
www.anthropic.com
November 3, 2025 at 7:18 PM
“Effective context engineering for AI agents” by Anthropic down best practices (Compaction, agentic memory, multi-agent) architectures) for composing high-quality context into components for LLM prompts for “just-in-time” loading rather than bulk everything upfront www.anthropic.com/engineering/...
"DeepSeek‑OCR" proposes way to ingest very large document corpora (e.g., patient histories, full trial protocols, longitudinal datasets) for LLM/VLM. This compression may be promising direction to reduce token count/compute while retaining fidelity but need to test accuracy arxiv.org/abs/2510.18234
DeepSeek-OCR: Contexts Optical Compression
We present DeepSeek-OCR as an initial investigation into the feasibility of compressing long contexts via optical 2D mapping. DeepSeek-OCR consists of two components: DeepEncoder and DeepSeek3B-MoE-A5...
arxiv.org
November 3, 2025 at 7:13 PM
"DeepSeek‑OCR" proposes way to ingest very large document corpora (e.g., patient histories, full trial protocols, longitudinal datasets) for LLM/VLM. This compression may be promising direction to reduce token count/compute while retaining fidelity but need to test accuracy arxiv.org/abs/2510.18234
“The Medium is the Message” reveals how patient message is presented can sway the outputs of clinical LLMs, potentially harmful ways. points to brittleness in LLMs when faced with real-world, non-clinical variation in inputs, with fairness implications (especially gender) dl.acm.org/doi/10.1145/...
The Medium is the Message: How Non-Clinical Information Shapes Clinical Decisions in LLMs | Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency
dl.acm.org
November 3, 2025 at 7:07 PM
“The Medium is the Message” reveals how patient message is presented can sway the outputs of clinical LLMs, potentially harmful ways. points to brittleness in LLMs when faced with real-world, non-clinical variation in inputs, with fairness implications (especially gender) dl.acm.org/doi/10.1145/...