Stanford/ARC Institute used AI to generate functional novel viruses (bacteriophages) for the first time. While currently limited to bacteria, the thousand-fold cost reduction in viral desig…

Stanford/ARC Institute used AI to design novel bacteriophages; while not human pathogens, the technique could accelerate viral vector engineering for gene therapies, but 'virus' branding ne…

Stanford/ARC Institute used AI to generate functional bacteriophages never seen in nature. While current viruses only target bacteria, cost collapse in viral design reshapes biotech positiv…

ARC Institute's AI-generated bacteriophages demonstrate path to novel viral delivery vehicles for gene therapy. Commercialization requires shedding 'virus' stigma post-COVID — analogous to…

OpenAI's hockey puck device with cameras/sensors/expressive movement represents first major lab bet on dedicated hardware. Simultaneously, Google's TPU sales to competitors (20%+ of shipmen…

OpenAI's first consumer device is a screenless, expressive, context-aware AI companion (hockey puck donut, $300-400) designed by Jony Ive. Targets 2027 launch as first in a device family to…

OpenAI's first hardware product includes cameras, sensors, moving parts, and expressive lights to create an always-present ambient AI assistant; success depends on integrating voice mode, m…

States are winning nine-figure judgments against Meta for youth addiction (NM: $900M+, KY: $9M). Hosts map the exact tobacco 1998 template: state-led suits over healthcare externalities → p…

New Mexico's $900M+ Meta judgment follows Kentucky school district case — pattern replicates 1998 Tobacco MSA where states secured indefinite annual payments financialized via bonds. Meta's…

Frontier labs live or die by compute allocation. Google directing >20% of TPU capacity to Anthropic instead of DeepMind reveals strategic prioritization of cloud revenue over model leadersh…

Google Cloud's decision to sell massive TPU capacity to Anthropic (a DeepMind rival) while starving internal research reflects strategic confusion. Compute allocation is the binding constra…

ABG monetizes iconic brand IP (Reebok, Brooks Brothers, Elvis, Marilyn Monroe, Tapout) via licensing rather than operating. Critics argue it's a 'graveyard' where brands stagnate without in…

Fund strategy rejects neo-lab acqui-hire bets and default pro-rata. Instead: incubate hard tech (fusion, space), take 20-30% ownership early, then either 3x pro-rata or 1/3 pro-rata — never…

Samir argues virus/hacking incidents show why US companies must stay at the cutting edge to understand and defend against threats; overregulation cedes advantage to adversaries with no guar…

Coastal Ventures argues billion-dollar seed rounds in 'neolabs' produce aquihires where investors lose (Windsurf, Scale AI examples). Sustainable venture returns require 20%+ early ownershi…

Coastal Ventures argues sustainable returns come from early, high-ownership bets (e.g., Rocket Lab at $5M for 33%), not aquihire-driven late-stage rounds where investors don't recover headl…

Coastal Ventures' incubation of Commonwealth Fusion demonstrates fusion's transition from science project to investable hard tech. Technical edge (high-temp superconductors) + regulatory ta…

Coastal Ventures' early incubation of Commonwealth Fusion Systems exemplifies their 'bold, early, impactful' hard-tech strategy: technical edge, company-building support, decade-plus horizo…

Coastal's 28% ownership at IPO from obscure NZ startup validates space launch as category where early technical bets yield fund returns. Growing launch demand (constellations, defense, AI d…

Enterprises adopting AI coding agents (Cursor, Codex, etc.) face unbounded consumption costs growing exponentially despite 2x productivity gains. Databricks built Unity AI gateway: smart ro…

Databricks' internal rollout shows coding agents double developer output, but unbounded token consumption requires active model routing (30% savings), rapid adoption of efficient new models…

Databricks' internal deployment across 10,000+ employees proves AI coding agents are the highest-ROI GenAI use case. Unity AI Gateway's dynamic routing across open/closed models maintains q…

Whatnot's transaction-native live video enables creators to earn $500K run rate in 90 days (75% of $1M+ earners) with only 15-20 concurrent viewers — economics equivalent to physical retail…

Whatnot's live commerce model monetizes an order of magnitude better than ad-supported platforms; sellers need only 15-20 concurrent viewers for viable economics, creating a broad middle cl…

Live video commerce monetizes an order of magnitude better than ad-supported platforms because transactions are direct. Power law exists but middle class thrives: consistent sellers do $100…