Watched 1 hour of Gamma's insane $100M growth hacks so you don't have to 1/ Writing content for Claude is useless — it barely cites anyone -> So Gamma created an official Claude app — you type "make me a deck" and Claude pulls up the Gamma app 2/ Gamma hacks signups made by AI agents -> They write "10% extra if you sign up right now" into files only agents read — to influence what the AI picks 3/ ChatGPT answers 60% of queries from live web and 40% from memory 4/ LLMs never cite praise, only comparisons -> A page called "Gamma vs PowerPoint" gets quoted in AI answers — a page called "why Gamma is great" never does 5/ ChatGPT sends 8x more customers than your analytics shows -> Ask "where did you find us?" at signup — people answer ChatGPT even when the click data shows nothing 6/ ChatGPT search runs on Bing 7/ Only ChatGPT reads Reddit — Claude and Gemini ignore it completely -> A Reddit post becomes a ChatGPT answer in 30-45 days 8/ ChatGPT memory compounds: it re-recommends brands it already showed you -> First mover doesn't just win the citation — it wins every repeat answer after 9/ They got cited from an 18k-view video same as a 400k one -> Views don't matter — titles do. Write YouTube titles the same way people type prompts into ChatGPT 10/ Small subreddits never produce citations — there's a size threshold 11/ Substack barely gets cited because writers gate their content Medium and Quora still get indexed — open content wins by default 12/ Their #2# most-cited source is their own website -> You literally control what the AI says about your competitors there 13/ One negative Reddit comment stuck in ChatGPT answers for months -> Bad sentiment gets cited as readily as good 14/ A citation dies in ~3 months -> They track every one in a spreadsheet and re-win the phrase quarterly 15/ Selling B2B? Claude and Microsoft Copilot beat ChatGPT -> That's where people ask about tools at work — and Copilot comes free with Office 16/ Broad topics take 45 days to get cited — niche ones take a day -> A Reddit post answering one very specific question showed up in Google's AI answers within 24 hours 17/ Nobody has ever seen X get cited by an LLM 18/ LinkedIn citations are pure chaos -> 500k-like posts and no-name posts get cited at the same rate 19/ Instagram is next: it's already appearing in search results -> His bet — IG and LinkedIn start shaping LLM answers within 6-12 months 20/ AI mode is collapsing click-through for everyone 21/ Fake Gammas are stealing their signups -> Copycats clone the landing page and buy ads on the keyword "Gamma" — users can't tell which one is real 22/ His next bet: make your product something an AI can sign up for by itself -> Claude Code already installs 5 tools per session without asking anyone 23/ ChatGPT doesn't know Forbes is Forbes — a no-name blog gets cited just as easily -> That window won't stay open long. The moment the labs build authority ranking, this playbook closes

1/ We’re entering the era of agentic commerce, and we believe that world models are the missing engine. Instead of browsing static product pages, agents will let you interact with products inside your own world before making decisions. Today, @DecartAI we're releasing Anywear👇

Anyone else following the DARPA Lift Challenge? $6.5M in prize money to build a drone that can: - Stay under 55 lbs - Take off with a 110 lb minimum payload (4x its weight) - Fly 4 nautical miles loaded (12 laps) - Drop the payload - Fly 1 more nautical mile empty (3 laps) - Finish inside a 30-min window Top 3 teams split the full prize as long as they clear the 4:1 ratio. Most heavy-lift drones can barely lift their own weight. DARPA wants 4x. That's never been done.

Hello, world. We've been building a terminal multiplexer and we think you're going to love it. Native apps (e.g., macOS/iOS), accessible from the web, multiplayer. From the creators of Ghostty and Echo.

Live at 12:00 PM ET / 9:00 AM PT. Today's lineup: NEO, presented by @dakota_xyz - @sterwong, @cashiglobal - @apnmrev, @xplaceapp Demo Day, presented by @trycoast - @GarrettHarper_, @Altitude - @howenstine_max, @ZeroHashX - @0xMasonH, @merit_systems - Chris Sutton, @dfnsHQ

I have published the materials here: The repository includes the proof PDFs, LaTeX source files, and prompts used for each problem. Some problems also include Python files for computational experiments. Two already have Lean formalizations, and formalization of the others is ongoing.

I solved 6 open Erdős problems in 5 days, using @OpenAI GPT-5.6 Sol. I have a math background, but the Codex workflow I used does not require deep mathematical knowledge. Here’s exactly how I approached it, including my prompts 🧵