Gavin Nelson

@ianzelbo @CharlesPattson Oh, you mean in their model they made the section of the glass black?

@ianzelbo @CharlesPattson What's the proper way to do this?

Everything about this absolutely slaps

@flyosity @MercedesBenz and piano black 🫠

While I was getting my car serviced, Mercedes let me borrow a new S 580e with a $144K sticker price The steering wheel has all touch controls made from cheap plastic while my 2022 GLE has aluminum buttons and knurled thumb controls What are we even doing here

@benjitaylor @elonmusk @nikitabier 🎉 let's go Benji!

I’m honoured to be joining 𝕏 to lead design. I believe this is the most important platform in the world, and I can’t think of a more exciting place to help shape the future. I’m looking forward to working closely with @elonmusk, @nikitabier, and the rest of the team. I’m grateful for the opportunity, humbled to be part of it, and can't wait to get started!

@RhysSullivan It should say when you are in thinking/pro right in the composer on web, no?

@samhenrigold @linear It's Robb's VisualizeTouches. Very helpful for recording videos of UI interaction!

While testing WireView, I noticed something in the @Linear app I hadn’t seen anywhere else: it detects if the screen is being captured or mirrored in some way and it automatically shows touch dots. Even supports multitouch! Could be a useful debug thing to add to your own apps

@heyimgustavo @sama Fix is coming out next release!

@Gavmn @sama 🫶🏻

pls fix @sama

@heyimgustavo @sama Sent this to the team!

@sama @Gavmn

@williambout 😍

Thuma just released a cute little desk thuma.co/products/studi…

@AndrewCurran_ > this model has a beautiful mind 💯% this is exactly the phrase I was trying to find to describe it

GPT-5.4 seems like an improvement in general intelligence. Most benchmarks are not really that important to me anymore, but this model has a beautiful mind. It sees further. Also, a noticeable improvement in prose, it is much better at turning a phrase conversationally. If you want to test its growth, go back to any previous session where you felt like you were working at a high level, use the 'branch in new chat' button, and ask 5.4 to take a look. It will usually see a little further along whatever path you were on. Even if you thought things were concluded. It's usually something small, but depending on what you are working on, small can be big! Even if it is small, it's usually something you didn't think of at the time, and seeing a new direction can be all it takes for the lightning bolt. The models can't make a big jump like that yet, but they can sometimes make small jumps in crazy directions. A human standing on that will be able to leap very far sometimes, I bet. I think for this type of collaborative work to work best, you need to give the model access to your chain of thought - this is general advice for any field. Include everything that you considered while writing your next response to the model. Get it all in the context window, particularly the dead ends that you ran down while thinking about what the model just said. You are probably never going to think in that direction again, but as long as it's in the window, the model sees it every exchange. Then one day it says, 'Hey, remember these three things you thought of that went nowhere? Well, the version of me from six months ago didn't notice anything at the time, but taken together they might mean…' And the models can't do the big explosion that goes off in your head at that point, not yet anyway, but I think increasingly they will help bring humans to that point. A reversal of current trends: the best part of the job is left to the human - to experience revelation.

@JonRoth @nikitabier If I'm engaged enough to tap through and set filters, maybe that's not true for me?

@Gavmn @nikitabier I imagine this is because users engage with more content in the for you algorithm than the filtered one, so they try to get you back into it ASAP

@nikitabier should the For You filter be sticky? I find I have to reset it to my preferences frequently

i'm so excited for what comes next

@flyosity So sorry Mike 😞

Last Sunday morning I was on FaceTime with my Mom as the kids excitedly told her about their upcoming school and sports activities This Sunday morning I'll be giving the eulogy at her funeral Tomorrow is not promised 💔 Love you Mom

@spottedinprod @OpenAI :)

chatgpt · tactile blob by @openai

@soleio @firstofkind @southpkcommons @GeometryofLuck Happy birthday Soleio!

Today is my birthday. I’m pleased with what got accomplished last year: debuting @firstofkind, 14 new portcos, @southpkcommons design chats, unveiling @GeometryofLuck But this story is the project I’m happiest with—the most personal, unexpected one: keeperfable.com

@imagecrate Cool—I wonder if it would feel more natural if you dragged the prompt itself rather than the stop generating control

Been working on what steering could feel like in ai chats. Less time spent editing, more drag + drop to keep it on track

@AdamKPx Why not just retweet the video

Credits to @Gavmn

iOS split view concept 🤌

@adrianmg @ryhanhassan

Vibe coding with the OG Framer in coffeescript is probably 🔥

@ChristianSelig Battery is a non issue but the USB C port is so sharp on my structural support pinky

Battery has surprisingly been a complete non-issue