nil

i turned my codebase into a city so that i can stop squinting at terminals. now i just watch my ai agents fly between buildings. and you can try it too: SQUAD: a fun mission control for ai agents

@random_ai_guy_ @marmaduke091 @JustinGorya w

@marmaduke091 @JustinGorya Exactly

claude-opus-4.8 is on Google Vertex 👀

@fabianstelzer @heyglif

Get automated *Motion Design* captions for your videos in @heyglif! Just upload your video and ask it to create dynamic captions in any style you want. Glif analyzes your video + cleverly adds fitting captions & *visualizations* based on everything *you're saying*!

@Cointelegraph @pbshgthm 😁

🇨🇳 NEW: Chinese cities are rolling out AI-powered robot barber kiosks that scan customers in 3D and cut hair with millimeter precision for just 60 yen per session.

sometimes claude's thinking traces are so much more fun to read than it's answers!

We need more fun app icons like these

@thsottiaux i made a fun flashcard app from a park yesterday, thanks to codex mobile app!

Codex is for cosy Sunday evenings. Show me your cosy creations.

@rronak_ i love SUPERHOT! played it a lot during the pandemic, but quickly finished all the levels! i'd be v happy if ai agents would be good enough to make infinite dynamically generated maps! 🙃

Ever since we launched windsurf, one of my internal evals for coding agents has been recreating the game SUPERHOT, a puzzle/action game where time only moves when you move. It's the perfect test of tricky game mechanics, simple but beautiful art style, and balancing level design. I have a robust prompt instructing the agent to make SUPERCOLD, and attach a reference image, and also allow 3 follow-up prompts that points out any mechnical issues. When Sonnet 3.5 came out, it could barely generate the world at all. Sonnet 3.7 and you could move around in the world, but it was still hilariously unplayable. Now, Opus 4.7 and it's playable, but still doesn't look great. What's needed to get all the way there? 1. Visual taste: Given a reference image, models should be able to discern whether their final output matches the reference image style. Shading, character design, etc, models should be unsatisfied with their current output 2. Stronger computer use: Frontier models somewhat attempt this, but agent should be able to play the entire level themselves, and iterate from there 3. Systems design: Complex projects are not yet written with scalable systems design as an experienced engineer would write it. This hinders the ability to create complex projects with a lot of moving parts. We're rapidly accelerating AI progress, let's see where we land in 3 months!

@HeatherRasley 🙈😂

@dhvanil O no Nil 😆

the optimum amount of ai psychosis is non-zero

weekly touch grass day!

@richiemcilroy those were some fun days...

just sat here thinking back to 2021, when you had to get approval from OpenAI to use GPT-3 in prod

in most enterprise software companies today, the FDE is doing more product design than the product designer is.

love to see 5.5 xhigh manage 12 subagents without breaking a sweat!

@sahra_bn that's beautiful!! incidentally, also how i imagine fresh new leaves swaying in sunny spring wind might feel! 🙃

ha :) that is a very nice connection, i haven’t made it (at least consciously) i meant to draw what it feels like to dance to music that moves me, specifically a choreography i made. and that is inspired by how it feels to receive love and truth from “the source” and give it back

"Lost in the latent space" -- 50x50cm acrylic on canvas!

@sahra_bn so cool! looks like a deconstructed version of 👀 x.com/sahra_bn/statu…

@dhvanil wow so cool. this is what i did today :)

@jxnlco yes

Do you want more software or better software?

so instead of saying “funky cabins within 2 hour drive” i will have to keep filling out your patient intake form

@simon_ohler 🫡 with some help from Claude! 💛

@dhvanil That’s you???? Damn brother

Crazy declassified 1966 footage of an Unidentified Friendly Opus (UFO)