absorb.md

The Investor's Podcast Network

Chronological feed of everything captured from The Investor's Podcast Network.

paper / investorspodcast / Apr 12

Lifelong Learning for Robust Spinal Cord Segmentation and Morphometric Drift Monitoring

This study introduces a spinal cord segmentation model trained on a diverse multisite dataset, incorporating a lifelong learning framework to monitor morphometric drift. The drift monitoring employs an automatic GitHub Actions workflow, recording morphometric values as the model updates. This framework enables continuous assessment of model stability and accuracy for clinical applications, particularly in developing normative values for spinal cord morphometry.

spinal-cord-segmentationmorphometric-driftlifelong-learningbiomarkersneurological-diseasesmri-analysiscomputer-vision
“A new spinal cord segmentation model outperforms previous versions and pathology-specific models, particularly in challenging lumbar spinal cord cases.”
paper / investorspodcast / Apr 12

Machine Learning for Real-Time Analysis at LHC

Machine learning (ML) is becoming increasingly vital for real-time analysis (RTA) in High Energy Physics (HEP), particularly within the challenging trigger systems of Large Hadron Collider (LHC) experiments. This whitepaper, compiled by the SMARTHEP network, provides a high-level overview of specific RTA use cases, demonstrating the breadth of ML applications. It also emphasizes the importance of collaboration between the HEP community and industry by showcasing interdisciplinary RTA examples from industrial contexts.

high-energy-physicsmachine-learningreal-time-analysislhc-experimentsdata-processingtrigger-systemsindustrial-collaboration
“Machine learning techniques are increasingly used in high energy physics (HEP).”
paper / investorspodcast / Apr 12

RareCollab: An AI-driven Multi-modal Diagnostic Framework for Mendelian Disorders

RareCollab is an agentic diagnostic framework that integrates genomic data, transcriptomic sequencing, and phenotypic information to diagnose rare Mendelian disorders. It combines a quantitative Diagnostic Engine with Large Language Model-based specialist modules for high-resolution, interpretable assessments. This approach aims to reduce the diagnostic odyssey for patients by operationalizing multi-modal evidence for accurate and scalable rare disease diagnosis.

mendelian-disordersmulti-modal-diagnosisgenomic-datatranscriptomic-dataphenotype-analysislarge-language-modelsrare-disease
“Current diagnostic methods for Mendelian disorders frequently fail to provide definitive molecular diagnoses.”
paper / investorspodcast / Apr 12

ILC Technology Network Progress Toward Linear Collider Realization

The ILC Technology Network (ITN), established in 2022, is actively conducting engineering studies for the International Linear Collider (ILC). This network\

accelerator-physicshigh-energy-physicslinear-colliderparticle-physicsengineering-studiesinternational-collaboration
“The ILC Technology Network (ITN) was founded in 2022.”
youtube / investorspodcast / Apr 10

Hermès Sustains Elite Luxury Moat Through Scarcity and Family Control, Justifying Premium Valuations

Hermès maintains a durable competitive moat in ultra-luxury via deliberate scarcity, relationship-based sales gating (e.g., Birkin bags require years of prior purchases), and family-led craftsmanship, enabling 17% revenue CAGR despite declining store counts. This insulates it from economic cycles, outperforming peers like LVMH (trading at 30x vs. Hermès' 50x PE) and delivering Birkin bags that beat S&P 500 returns over decades. At 50x earnings, DCF models yield mid-single-digit returns unless growth reaccelerates to mid-teens, lacking margin of safety for new positions.

luxury-goodsequity-researchvaluation-analysisconsumer-behaviorbrand-strategy
“Hermès Birkin bags have outperformed the S&P 500 with 14% annualized returns over 3-4 decades”
youtube / investorspodcast / Apr 10

Netflix's Enduring Moat: Subscriber Stickiness, Pricing Power, and Streaming Dominance Despite Recent Pullback

Netflix maintains a superior competitive position through high engagement via its recommendation algorithm, low churn around 2.5%, and a flywheel of original content driving retention and word-of-mouth acquisition. The company has pivoted from subscriber growth to monetization via price hikes, ad tiers, and password crackdowns, achieving 30% operating margins and $7B free cash flow in 2024 with ARPU at $11.64 globally (up to $17.20 in US/Canada). International expansion with localized hits like Squid Game, proprietary CDN for reliability, and potential Warner Bros. acquisition of HBO/DC IP position it to capture more TV viewing share as cable declines, enabling double-digit EPS growth.

netflixstock-analysisstreaming-industryvalue-investingmergers-and-acquisitionsmedia-economics
“Netflix stock dropped 30% since last summer amid Warner Bros. bid interest”
Older entries →