Podbit · Lex Fridman Podcast
Compilers cannot auto-vectorize anywhere close to what an expert human can do by hand. The dav1d AV1 decoder proves it: 80% assembly, 62x faster than C on pixel conversion functions, running on 3 billion devices. Moore's Law is over — squeezing hardware through hand-crafted assembly is how you get more without buying more.
Where this was said
At 1:50:40 · chapter starts 1:48:59
Kostya's GoToMeeting reverse engineering, stepping through millions of CPU instructions, dumping memory from VMs, needing bit-exact output — and why a 1MB blob takes a month. [1] — Kieran Kunhya "Reverse engineering a codec means opening a binary with no documentation, stepping through millions of CPU instructions in a debugger, dump…" 2:00:00
Hand-written SIMD assembly for pixel format conversion functions in Kieran's company runs 62 times faster than equivalent C code.
AV1 achieves 40–60% bandwidth reduction compared to H.264 at the same visual quality, enabling better streaming at lower bitrates.
Reverse engineering a codec means opening a binary with no documentation, stepping through millions of CPU instructions in a debugger, dumping memory from a virtual machine, and slowly inferring — frame by frame — what the compression algorithm is doing. A 1MB blob takes roughly a month. Kostya did 20–30MB blobs regularly, for fun.
VLC's latest version still supports iOS 9 through the current iOS 26, achieved through elaborate Frankenstein compiler configurations.
FFmpeg's FATE testing system runs on volunteer hardware across hundreds of compiler, OS, and CPU variants — from Visual Studio to Apple Clang, from x86 to RISC-V to PowerPC. A single code change can silently miscompile on GCC 11 on macOS. The matrix of combinations isn't a table — it's a pivot table.
Game apps keep users engaged long enough for ads to pay off. Tool apps don't — so if you're building a utility, ads are almost always the wrong call and subscriptions are your only real lever.
Inside SEO Bot, a single button labelled 'Boost My Domain Rating' routes users directly to Listing Bot. That one interaction converts a user of one tool into a user of two — without any marketing cost.
Directory listings are a powerful but underrated growth channel — but only if your product is genuinely interesting enough to earn the click. AI products have a natural advantage here because they're easy to package in a compelling, clickable way.
The fastest path to Twitter growth isn't volume — it's forming sharp opinions about how the platform works and sharing them immediately. People cluster around those who understand the rules and say so out loud.
Sam had no coding knowledge, so he used ChatGPT voice mode to generate his entire codebase and copy-pasted it into Notepad. A friend later introduced him to Cursor, and he never looked back.
Copy days of Discord chat history, paste it into ChatGPT, and ask it to list recurring pain points. The ones that come up most often are your best product bets.
Sam's top advice: when prompting Cursor, tell it to architect code for 100,000 users from day one. The AI changes its approach, building scalable frameworks instead of brittle one-user code.
With AI coding tools like Cursor, Bhanu replicates an existing free tool for a new keyword in under 5 minutes. What used to be a multi-day build is now a lunch-break task.
Ahrefs, SiteGPT, Cal.com, PostHog, Datafast, Sibyl AI, Bento, Feather, Featurepace, Mintlify, Cloud Code, ChartMogul — Bhanu runs his entire business solo with these 12 tools.
We use essential and analytics cookies to run Vuci. To understand how the site is used: Privacy Policy.
Install Vuci on your phone
Add it to your home screen for a faster, app-like experience.
Install Vuci on your phone
Tap the Share button, then “Add to Home Screen”.
A new version is available
Reload to get the latest Vuci.