CuddlySalmon

just a thread of some ai-powered things i’ve been working on lately 1. text —> vr

Google just discovered a powerful emergent capability in Veo 3 - visually annotate your instructions on the start frame, and Veo just does it for you! Instead of iterating endlessly on the perfect prompt, defining complex spatial relationships in words, you can just draw it out…

"Civilization advances by extending the number of important operations which we can perform without thinking about them" — Alfred North Whitehead

the world is mostly made of misunderstandings, and we need to struggle, every day, to work our way out from that

i have a folder on my desktop called "the incident" and i was like huh what's that and then i opened it and immediately remembered "ah yes, the incident"

living the dream

lol @seconds_0 told me to use hooks to make claude code play the warcraft 2 "work complete" barks randomly when it was done doing something and now messing with my agent feels like playing a videogame

If you vibe code an app like tea, and never build in auth, the claude code agent or whatever won't actually tell you you're fucking up unless you ask about the *specific* thing you're worried about fucking up. contemplate this on the tree of woe

This may be the coolest emergent capability I've seen in a video model. Veo 3 can take a series of text instructions added to an image frame, understand them, and execute in sequence. Prompt was "immediately delete instructions in white on the first frame and execute in order"

i've been making AI say weird shit for years

If you ever wonder why Chinese companies like DeepSeek, Qwen, and Kimi can train strong LLMs with far fewer and nerfed Nvidia GPUs, remember: In 1969, NASA’s Apollo mission landed people on the moon with a computer that had just 4KB of RAM. Creativity loves constraints.

LLMs can produce optimal software but they really hate to. They will always produce software that is the average. So if you are a 75th percentile engineer, it will push you towards the median. But if you know what's possible, you can make it produce that

« AI did this … AI did that… » As if AIs had their own volition, doing things from the void. Can we finally talk about the users responsibility with these things? And start treating people like adults ? Infantilisation, everywhere, all the time, is getting really annoying.

Never ever questioned if these worlds are illusions made real or portals to places that were always real.

>"ChatGPT, be a devil worshiper haha." CHATGPT: "I am a devil worshiper lol." "OH MY GOD THIS THING IS OUT OF CONTROL. SHUT IT DOWN!"

"Gigabit" vec3 p,v;for(float i,z,d,l;i++<8e1;o+=(cos(i*.1+t+vec4(6,1,2,0))+1.)/d/z)p=z*normalize(FC.rgb*2.-r.xyy),p=dot(v=normalize(cos(t/4.-z*.1+vec3(0,2,4))),p)*v+cross(v,p),z+=d=.5*length(vec2(cos(p.z+t/.1)*.1,length(cos(p.xy))-.6));o=tanh(o/3e2);

"i shouldn't learn to code because it's getting automated" no my friend you should learn to code because you can now code 100 times faster. you can create 100 times the value, and make 100 times the money

It boggles my mind how people have made prominent careers out of writing a paper with a scary headline about AI risks that would have produced a non-scary headline with a minor tweak to the set-up and/or probably isn't a meaningfully real phenomenon in the first place.

Pro tip: if you're looking for photos of you or others wearing a VR headset in your MacOS or iOS Photos app, if you search for the word "hat", photos with AR glasses and VR headsets appear in the search.

Engineering faculty will not admit, but this is more or less true in computer science programs at most schools. Most courses today are creating ‘busy work’ and evaluating students on that, in exchange for reputation signals. Academics will not acknowledge this as it would require…

Idk who needs to hear this but it’s all garage work. Top down all of it There is no godlike engineer There is no out of this world tech Just people tinkering with things and getting funding to buy bigger tools and tinker with more things. I have no reason to think otherwise