reSee.it - Related Video Feed

Artificial intelligence is returning to medicine not as a curiosity but as a driver of drug discovery and development. The typical pipeline begins with a biological insight and a therapeutic hypothesis about how modulating a target could help a patient, then moves to creating the right chemical matter, and finally to clinical development in people. The farther you go, the more expensive it gets, with clinical development being the costliest and most failure-prone stage. Depending on the estimate you trust, only about 5 to 10 percent of molecules entering the final clinical phase emerge with regulatory approval. The industry’s cost mood has spiraled, with fully loaded programs now soaring north of 2.6 billion dollars. Advances in AI are accelerating the middle piece of this journey: turning a target into a drug by designing effective molecules and screening vast libraries. The protein space benefits especially because advances like AlphaFold give structural context that makes it easier to predict how a molecule will interact with a protein. In addition, the explosion of multi-modal biological data—from cells and tissues to single-cell profiling and imaging—creates raw material for AI to interrogate biology at scale. Yet there is a gap: AI can rapidly generate hypotheses and designs, but turning new biological insights into disease-modifying therapies remains the harder, slower part of the journey. The strongest potential lies in redefining disease biology itself and identifying precise subtypes that respond to specific interventions. Data and incentives shape what is possible. A transformation in health care data collection and sharing is needed: richer, harmonized data from patients, with appropriate anonymization and safeguards. The talk notes that incentives in the United States often do not align with comprehensive diagnostics and data-driven treatment choices, and that centralized health data repositories could unlock breakthroughs much faster. Collaboration between academia and industry is essential, balancing deep theoretical thinking with product-like execution. The optimism rests on an exponential trajectory across AI, biology, and medicines, with the pace of change accelerating as measurement improves and integration tightens, ultimately enabling more precise, effective therapies.

In this a16z podcast, Gabriel Otte, CEO of Freenome, discusses advancements in cancer detection using machine learning and genomic data. The Human Genome Project revealed the complexity of genetics, leading to a systems biology approach that integrates computational tools to analyze gene interactions. The cost of genomic sequencing has significantly decreased, enabling broader applications in diagnostics. Early cancer detection is crucial, as survival rates dramatically improve when diagnosed before symptoms appear. Current challenges include convincing insurance companies to reimburse genomic tests, which often have high costs and uncertain ROI. Future opportunities lie in diverse applications of genomics, including mental health and infectious diseases, while advancements in technologies like CRISPR and mass spectrometry hold promise for further breakthroughs.

Every frontier in medicine seems to orbit a paradox: the more we know, the more ambitious the questions become. Mukherjee describes a career trained in reverse—immunology first, then biology, then medicine—and explains how medicine and science form a single yin-and-yang approach. He traces influences from Paul Berg at Stanford to Alan Townsen in Oxford and a Harvard postdoc, using these threads to frame a view that cancer research spans prevention, early detection, and treatment. Prevention, he says, remains underfunded but yields the best returns, as researchers explore how inflammation, obesity, air pollution, and diet influence cancer risk. Early detection increasingly relies on AI and new screening strategies, while treatment mobilizes the body’s defenses and novel drugs. Mukherjee describes ImmunoACT, an effort to bring CAR T therapies to India. These therapies require extracting T cells, engineering them, and reintroducing them to attack cancer, with the aim of cutting costs and broadening access. About 25 patients have been treated, with cure data for certain leukemias and lymphomas matching U.S. outcomes, illustrating democratization. He sees AI as both diagnostic aid and driver of new medicines, including computationally designed drugs. He notes AlphaFold’s protein folding and argues the lock-and-key problem in drug design can be accelerated by generative AI. Mukherjee widens the frame to ask what humanity is becoming. He proposes a five-seat spaceship: a pure humanist philosopher, a historian, a pure scientist, a translator who bridges science and society, and a technologist-inventor who can deploy new capabilities; a second translator, a science-fiction writer, and a tribal leader would round out the crew. He treats AI as both opportunity and risk, urging creativity, empathy, and diversity while safeguarding gene–environment interactions. He contrasts disease and enhancement, arguing that culture and memes may precede genetics. References to IVF, bone marrow transplants, Pollock, Rhodes, and Billie Holiday anchor the discussion in resilience and imagination. The takeaway: align technology with human flourishing for a peaceful, creative future.

In this episode of the Drive podcast, host Peter Attia speaks with Max Diehn, a fellow Stanford MSTP graduate, about his journey through medical school, his research in cancer biology, and the development of liquid biopsies. Diehn shares his experience in Pat Brown's lab, where he worked on DNA microarrays, a revolutionary technology for measuring gene expression across the genome, which opened new avenues in immunology and oncology. Diehn discusses the challenges of working with RNA, particularly its instability, and the importance of preserving samples to minimize degradation. He highlights his dissertation projects, which focused on T cell activation and the cataloging of gene expression changes, contributing to the understanding of immunotherapy. Transitioning to clinical training, Diehn reflects on his decision to pursue radiation oncology, influenced by his father's battle with lymphoma and the desire to improve cancer treatments. He emphasizes the importance of balancing clinical practice with research, particularly through the Holman pathway during his residency, allowing him to maintain a foot in both worlds. The conversation shifts to liquid biopsies, a field Diehn became interested in due to the limitations of traditional imaging methods in detecting early-stage cancers. He explains the potential of liquid biopsies to identify circulating tumor DNA (ctDNA) and how it can provide insights into cancer recurrence and treatment efficacy. Diehn notes that while protein biomarkers like PSA and CEA have been used historically, they lack specificity and sensitivity compared to ctDNA. Diehn elaborates on the technical challenges of detecting ctDNA, emphasizing the need for sensitive methods like next-generation sequencing to identify mutations unique to cancer cells. He discusses the potential of combining mutation analysis with other factors, such as methylation patterns, to improve the accuracy of liquid biopsies for cancer screening. The discussion also touches on the current state of liquid biopsy research, including ongoing trials in lung, breast, and colorectal cancers. Diehn expresses optimism about the future of liquid biopsies, particularly in guiding adjuvant therapy for patients with early-stage cancers. He acknowledges the need for rigorous clinical trials to validate the effectiveness of these tests in reducing cancer-specific mortality. Attia and Diehn conclude by discussing the broader implications of liquid biopsies in cancer screening and the importance of developing reliable tests that can be used in clinical practice. Diehn emphasizes the need for continued research and collaboration to advance the field and improve patient outcomes.

In this episode of the Drive podcast, Peter Attia interviews Alex Aravanis, discussing their shared background in engineering and medicine, and their journey into the field of genomics and cancer diagnostics. Alex, who transitioned from electrical engineering to a PhD in neuroscience, highlights his work on synaptic communication in the brain, which laid the groundwork for applying engineering principles to biological questions. Alex joined Illumina in 2013, a company known for its DNA sequencing technologies. He was recruited to develop clinical applications of sequencing technology, particularly in oncology. He explains the evolution of DNA sequencing from the Human Genome Project to the current state, where sequencing costs have dramatically decreased from billions to a few hundred dollars, thanks to innovations like next-generation sequencing. The conversation shifts to liquid biopsies, which allow for the detection of tumor DNA in blood samples. Alex discusses the historical context of liquid biopsies, starting with tumor sequencing and the discovery of cell-free DNA in cancer patients. He explains how this technology can identify mutations in tumors without invasive biopsies, making it particularly valuable for late-stage cancer patients. Alex elaborates on the development of Grail, a company focused on early cancer detection through liquid biopsies. The discussion covers the importance of specificity and sensitivity in cancer screening tests, emphasizing the need for high specificity to avoid unnecessary follow-ups for false positives. Alex shares insights from Grail's studies, including the Gallery test, which aims to detect multiple cancers at once and has shown promising results in identifying cancers that are often missed by traditional screening methods. The podcast also explores the role of methylation in the epigenome and its potential implications for aging and cancer. Alex discusses the significance of methylation patterns in determining cell identity and how they can be used to predict the origin of cancers detected through liquid biopsies. He highlights the need for further research into the epigenome and its relationship to aging, suggesting that understanding and potentially reversing epigenetic changes could lead to rejuvenation therapies. The conversation concludes with a forward-looking perspective on the future of cancer detection and treatment, emphasizing the potential for advancements in technology and biology to improve health outcomes and extend healthspan. Alex expresses optimism about the possibilities for rejuvenation therapies targeting specific tissues and organs, driven by ongoing research in epigenetics and cell biology.

An audacious story unfolds from a Stanford professor who braids cancer biology with a data revolution. He describes the immune system’s daily dance with tumors, where mutations drive cancer, tumors learn to turn off MHC presentation, and the immune system can be misled into helping cancer spread. He personalizes this with his own melanoma and kidney cancer linked to a MIDF 318K mutation, revealed by genome sequencing. Early detection remains central, and he emphasizes that the immune system governs every stage—from precancerous lesions to metastasis—shaping how therapies are chosen and timed. He then explains the breakthrough role of immune checkpoint therapy, referencing Jim Allison’s Nobel Prize and trials that showed 5% survival in melanoma rising to about 50% when the immune brake was released. The discussion covers how tumors initiate disease, evade surveillance by mutating antigen presentation, and how drugs and diagnostics aim to restore immune recognition. The guest describes the progression from benign lesions to metastatic cancer as a multi-step race, where reactivating the immune system at the right moment can prevent spread and tailor treatment to each patient’s tumor subtype. Beyond biology, the guest describes a data revolution in immunology. He explains how his lab built instruments to measure 50–60 proteins at once, enabling near-complete mapping of immune-cell types and their roles in cancer. The data feeds mathematical models and pseudotime analyses that illuminate the paths from normal cells to leukemia, and they underpin efforts to personalize medicines. He notes that his work helped spark a suite of companies, including a project that sold to 10x Genomics, and he emphasizes the need to fuse diagnostics with targeted therapies to improve outcomes. The conversation also dives into UAPs, M-shaped metals, and the promise of new instrumentation. The guest recounts sequencing the Otakama mummy as human and Chilean, and describes other meteorically unusual materials—silicon with magnesium isotope ratios suggesting neutron exposure contexts—and cases like the Council Bluffs molten-metal find. He argues for careful, peer-reviewed analysis, open data versus secrecy, and the potential for public–private partnerships to study artifacts without circus-style media. He discusses Skywatcher, Havana syndrome, and DoD interest, while imagining atomic-imaging tools that could map materials at the atomic level and accelerate discovery across science and medicine.

Jo Bhakdi, founder of Quantgene, shares his journey from a bioscience background to creating cutting-edge cancer detection technology. Growing up in a family of scientists, he initially pursued economics before returning to the life sciences after tackling a complex genomics problem in 2014. He emphasizes the importance of early cancer detection, noting that 600,000 people die from cancer annually in the U.S. alone, and highlights the stark survival differences between early and late-stage diagnoses. Quantgene focuses on deep genomic sequencing, which allows for the identification of cancerous mutations in blood samples. This technology surpasses traditional methods by analyzing all DNA fragments rather than a limited sample, significantly increasing accuracy. Bhakdi explains that while hereditary testing identifies predispositions to cancer, somatic testing reveals current tumors, making it more actionable. He discusses the challenges of integrating advanced technology into the medical field, particularly the need for a self-payer model to drive innovation. By charging a membership fee for ongoing genomic testing and insights, Quantgene aims to align patient interests with healthcare advancements. Bhakdi also touches on the potential of gene editing technologies like CRISPR, emphasizing the need for extensive clinical trials to ensure safety. Ultimately, he advocates for a paradigm shift in medicine towards a data-driven, dynamic approach that prioritizes patient outcomes.

In this conversation, Regina Barzilay, a professor at MIT and a leading researcher in natural language processing and deep learning applications in oncology, discusses her insights on science, literature, and personal experiences. She emphasizes the importance of books in shaping her worldview, particularly highlighting "The Emperor of All Maladies," which reshaped her understanding of the scientific process and the complexities of cancer treatment. Barzilay reflects on how personal experiences, such as her own breast cancer diagnosis in 2014, shifted her perspective on the significance of scientific work, urging a focus on alleviating real-world suffering rather than trivial academic pursuits. She discusses the role of machine learning in early cancer detection, noting that many cancers are diagnosed too late for effective treatment. Barzilay believes that advancements in AI could lead to earlier detection and better utilization of existing treatments, although she expresses concern about the slow pace of medical establishment adaptation. She highlights the challenges in accessing large datasets for research, emphasizing the need for better data-sharing mechanisms and addressing privacy concerns. Barzilay also touches on the future of drug design, suggesting that machine learning could revolutionize the field by predicting molecular properties and generating new compounds. She encourages students interested in machine learning to find meaningful areas to apply their skills and stresses the importance of understanding the broader implications of their work. Ultimately, she advocates for a balance between personal mission and societal impact, urging individuals to remain true to their values amidst external pressures.

In this episode of Moonshots, Peter Diamandis speaks with Naveen Jain, CEO of Viome, and Guru Banavar, CTO and head of AI at Viome, about the intersection of artificial intelligence (AI) and healthcare. Jain emphasizes the need to ask different questions to tackle massive health problems, particularly chronic diseases, which account for 97% of healthcare spending. He highlights the importance of understanding the human microbiome, stating that 99% of genes expressed in our bodies come from microbes rather than human DNA. This insight shifts the focus from traditional genetic analysis to understanding RNA and microbial interactions. Viome aims to digitize human biology by collecting extensive biological data, including one quadrillion data points from the oral microbiome alone. Jain explains that the healthcare system has historically neglected the microbiome, treating it as a threat rather than a partner in health. By utilizing AI, Viome analyzes this vast data to identify patterns and correlations that can inform personalized health recommendations. Guru Banavar discusses the evolution of data processing capabilities, noting that recent advancements in computational power and algorithms have made it possible to analyze biological data at unprecedented scales. This allows for a deeper understanding of individual health and the development of personalized interventions. Jain outlines Viome's moonshot goal: to prevent and reverse chronic diseases through personalized nutrition, viewing food as medicine. He shares the company's journey, including the acquisition of RNA analysis technology from national labs and the development of consumer products that provide tailored health insights. The conversation also touches on the future of healthcare, predicting a shift towards preventative measures and the democratization of health information. Jain and Banavar envision a future where AI-driven tools provide real-time health guidance, enabling individuals to take control of their well-being. The episode concludes with a discussion on the importance of continuous learning and adaptation in health management, emphasizing that personalized approaches are essential for effective treatment. Jain encourages listeners to explore Viome's offerings to better understand their health and optimize their microbiome.

In this episode, Lex Fridman speaks with Manolis Kellis, a professor at MIT and head of the MIT Computational Biology Group, focusing on the complexities of human disease, genetics, and biology. Kellis emphasizes that understanding human disease is one of the most complex challenges in modern science, as it intertwines with the complexities of the human genome, brain circuitry, and various biological systems. Traditionally, research began with model organisms to understand basic biology before applying findings to humans. However, Kellis notes a paradigm shift where human genetics now drives basic biology, with more genetic mutation information available in the human genome than in any other species. He discusses the importance of perturbations—experimental manipulations to understand biological systems—and how genetic epidemiology correlates genomic changes with phenotypic differences, allowing researchers to identify disease mechanisms. Kellis explains that every individual carries approximately six million unique genetic variants, which can be viewed as natural experiments. This genetic diversity complicates the understanding of disease mechanisms in humans compared to simpler animal models. He highlights the significance of identifying disease pathways and understanding how specific genes relate to diseases, which can lead to targeted interventions and lifestyle changes. The conversation touches on the importance of understanding diseases like heart disease, cancer, and Alzheimer's, emphasizing their impact on quality of life and mortality rates. Kellis discusses the role of genetics in these diseases, noting that while some conditions have strong genetic components, environmental factors also play crucial roles. For instance, Alzheimer's has a significant genetic basis, but lifestyle changes can still influence its onset. Kellis elaborates on the advancements in technology that enable researchers to analyze genetic data at unprecedented scales, including single-cell RNA sequencing and CRISPR gene editing. He describes how these tools allow for the exploration of complex biological questions, such as the interactions between different cell types in the brain and their implications for diseases like Alzheimer's and schizophrenia. The discussion also covers the need for interdisciplinary collaboration, as understanding the circuitry of diseases requires insights from various fields, including immunology, neurology, and metabolism. Kellis argues for a systems medicine approach, where interventions target networks of genes and pathways rather than individual genes, leading to more effective treatments. Kellis concludes by expressing optimism about the future of disease research and treatment, highlighting the potential for new technologies and insights to revolutionize our understanding of health and disease. He envisions a future where personalized medicine can effectively address the complexities of human biology, ultimately improving health outcomes across populations.