a=(x,y,d=mag(k=9*cos(x/8),e=y/8-12.5)**2/99+sin(t)/6+.5)=>point((q=99-e*sin(atan2(k,e)*7)/d+k*(3+cos(d*d-t)*2))*sin(c=d/2+e/69-t/16)+200,(q+19*d)*cos(c)+200) t=0,draw=$=>{t||createCanvas(w=400,w);background(9).stroke(w,66);for(t+=PI/45,i=1e4;i--;)a(i%200,i/55)}#つぶやきProcessing

If you're frustrated at work, consider joining a smaller company

ChatGPT's growth youtu.be/GhIJs4zbH0o?si…

How beautiful.

.sarah guo // conviction at AI Engineer World's Fair highlighting Token Marketshare by Model Author. Tbh I'm a little surprised to see Google roughly twice that of OpenAI 🤔

Congrats to AI Engineer 2025 Best Speakers! MCP: David Cramer Tiny Teams: alex duffy LLM Recsys: Devansh GraphRAG: Daniel Chalef Fortune 500 Day 1: Harrison Chase Architects Day 1: Denys Linkov Infra: Dylan Patel Voice: Brooke Hopkins Product Management: Brian Balfour Agent

How do reasoning models solve hard math problems? We asked 14 mathematicians to review o3-mini-high’s raw, unsummarized reasoning traces on 29 FrontierMath problems. Here’s what they found:

AI capabilities are doubling every 70 days for coding complexity. Every 7 months for task length. That's 16-64x capability increase per year. We're not in gradual improvement territory anymore. (h/t Scott Wu from Cognition)

Extract – a system built by the UK government, using our Gemini foundational model – will help council planners make faster decisions. 🚀 Using multimodal reasoning, it turns complex planning documents – even handwritten notes and blurry maps – into digital data in just 40s.

London VC by station dealroom.co/london-tech-we…

ChatGPT's product retention curves is a product manager's wet dream. Their 1 month retention has skyrocketed from <60% 2yrs ago to an unprecedented ~90%! Youtube was best-in-class with ~85%. 6mo retention is trending to ~80%. Rapidly rising smile curve. Generational product.

every single cluster of these is a viable startup btw my contribution to sarah guo // conviction’ excellent @aidotengineer talk about how to become a gpt wrapper millionaire is “take something people are ALREADY doing in raw chat and pave the cowpath” in every million copy pastes to/from

Link: blog.nilenso.com/blog/2025/05/2… Other great reads: nicholas.carlini.com/writing/2024/h… anthropic.com/engineering/cl… simonwillison.net/2025/Mar/11/us… crawshaw.io/blog/programmi…

The five types of product managers according to Peter Deng (ex-OpenAI, @Instagram, Uber, @Facebook, @Airtable): 1. The consumer PM: Half designer, half product person. Obsessed over the details, asking questions like "Is it delightful?" and "Is it crafted well enough?" 2. The

.Replit ⠕’s ARR since founding The space is … beyond on fire No one is going to code without AI again

+1 for "context engineering" over "prompt engineering". People associate prompts with short task descriptions you'd give an LLM in your day-to-day use. When in every industrial-strength LLM app, context engineering is the delicate art and science of filling the context window

Consumer ChatGPT & Google Search Global Desktop User Retention Rates via Benedict Evans ben-evans.com/benedictevans/…

Ongoing price collapse for OpenAI models via Benedict Evans ben-evans.com/presentations

Gallup did a poll highlighting that AI adoption is growing rapidly amongst US workers. But only 8% are using AI daily. And I bet that most people are only using ChatGPT. So I think there's still massive upside for employees who learn how to use the best AI tools to scale their

Great read on "context engineering" for agents rlancemartin.github.io/2025/06/23/con…