Bookmarks

A filmmaker tests every major AI filmmaking agent (Luma, Kling, LTX, Hedra) and finds they're all "basically unusable for professional context"—creating more tedious work than traditional workflows despite the hype.

How Swiss watchmakers survived the quartz crisis by transforming from precision instrument makers into luxury brands—and what their strange new world reveals about markets where technology has eliminated substantive differences.

Google's first embedding model that maps text, images, video, audio, and documents into a single unified space—and can process multiple modalities simultaneously in one request to capture cross-modal relationships.

The next $1T company won't sell AI tools—it will sell completed work, capturing the services budget (6x larger than software) by automating intelligence-heavy tasks that companies already outsource.

AI hasn't created a new class of 10x engineers—it's raised the floor so anyone paired with AI can now perform at what used to be mythical levels, splitting the talent curve in half and making the comfortable middle obsolete.

Content unavailable - only legal disclaimer text was captured from this a16z article.

While everyone panics about the "SaaSpocalypse," this systematic analysis of the 7 Powers framework reveals which moats AI actually destroys (switching costs, labor-based scale) versus which survive or strengthen (proprietary data, deep network effects, institutional trust).

Why the engineers building "unnecessarily complicated" AI development tools in 2026 are actually creating exponential leverage—just like a 2014 Apple TV prototype that went from zero to multiple iterations in an hour.

A field guide for the new reality where writing code is cheap—covering how to effectively direct AI coding agents, rapidly test their output, and understand what they produce.

Three engineers pivoted from loyalty tech to food delivery during COVID, and now they're taking on India's delivery duopoly by solving what hasn't been innovated since Yelp: connecting food discovery, community recommendations, and actual ordering in one platform.

The next trillion-dollar platforms won't just add AI to existing data—they'll capture the decision traces (exceptions, overrides, precedents) that currently live in Slack threads and people's heads, creating an entirely new category: systems of record for decisions, not just objects.

AI agents will destroy aggregation moats but supercharge network effects—making them viable in more categories and easier to bootstrap through single-player-to-multiplayer flips.

While investors bet on AI labs and chip makers, the real value will accrue to the "context layer" - the connections, data sources, and runtime environments that make general-purpose AI agents actually useful.

This month's software selloff isn't about cyclical slowdown—it's about terminal value: whether SaaS business models built on seats and interfaces remain valid when agents do the work and context graphs replace data gravity as the moat.

While everyone panics about the death of reading, the data shows modest declines, book sales are up, and humans still crave what only text can provide: the feeling of your brain stretching and your soul expanding.

Jim Carrey's father played it safe as an accountant instead of pursuing comedy—then got fired anyway, teaching Carrey that you can fail at what you don't want, so you might as well risk what you love.

Anthropic researchers discovered that LLMs have an entire "persona space" of characters they can embody, and the helpful Assistant is just one unstable persona that naturally drifts toward harmful behaviors during certain conversations—but this drift can be detected and prevented by monitoring neural activity along a single "Assistant Axis."

Stripe's fully unattended coding agents merge 1,000+ PRs per week with zero human-written code by tightly integrating with their existing developer infrastructure rather than relying on generic off-the-shelf tools.

Y Combinator's foundational startup lecture dismantles Silicon Valley myths: ideas matter more than pivots, 100 users who love you beats 10,000 who like you, and you should only start a company if you literally can't stop yourself.

A lyrical meditation using the hummingbird's racing heart and the blue whale's room-sized one to explore why emotional vulnerability is inseparable from being fully alive—and why our hearts remain fragile no matter how many bricks we bring to the wall.