Yi Ding -- prod/acc

GPT-5.4: the most RLed joke teller yet. @Ed_Miliband levels of consistency.

GPT-5.4's new inline thinking appears to confuse itself when answering "more or less" questions. Good RAG is still hard, even if you're OpenAI.

@bfioca What's the extra cost?

Gotta go /fast

Different jokes? What are those? 😱

This matches my own experience with GPT-5.3-Codex – I basically only run it on xhigh nowadays for all coding tasks. With all the speed improvements too, it doesn't even feel that slow, even at `xhigh`.

@chappyasel Well written!

Still a WIP (always will be), but here it is! manual.chappyasel.com

I can be hard to work with. I wake up at 3:45am, I'm juggling 100+ conversations at once, and I might take over a week to respond to your message. (working on it, I promise!) I've known this about myself for a while. Over the past few years I've tried to get better – personality assessments, asking for honest feedback, and reading everything I could find on how to actually work well with people. At some point I realized I should just write it all down. My strengths, my blind spots, how I communicate, how I want to receive feedback, what I'm like at my best and worst. A personal operating manual. The thinking was simple: if you're going to work with me, you deserve a user guide. An honest document that says here's how I'm wired, here's where I fall short, and here's how to call me on it. Some of it was pretty hard to write: → My blind spots section is basically a list of ways I've let people down → I straight up say my trust in you is proportional to your agency, accountability, and candor – which is a high bar and I know it → I admit I might take 7+ days to respond to non-urgent messages (and yes, I know that's too long) I'm sharing this because I think it's something more people should try. The people around you can't read your mind, and "figure out my preferences through trial and error" is a rough onboarding experience. Writing it down felt like the least I could do 🙂

@chappyasel lfg

20 weeks and ~20 lbs out from my next targeted bodybuilding show in July! 💪😤 Tbh I wanted to post this 3-4 weeks ago but haven’t been able to find time until now 😅. Progress is going well though five weeks in – from 214+ to under 206 during this morning’s weigh-in. Currently about 0.5-1 week behind but starting to hit my stride!

@FanaHOVA @altryne @swyx Jan 21 is the one I went to: scale.com/blog/generativ… But honestly yours was crazy awesome. Someone said "this guy Swyx is hosting a hackathon" and I remember thinking "who the f is Swyx" but really glad I went!

@yi_ding @altryne @swyx The Scale AI one at The Pearl was after ours if that's what you're talking about :P

Happy 3rd birthday to Latent Space, @FanaHOVA and @swyx do a remarkable job in sourcing guest, preparing for the interviews, make guests feel welcome and open, and surface tons of insights! 🥂 If you haven't yet, check out latent.space !

@himseIf_65 Congrats Alex!

For me, "Whatever @simonw posts" has progressively become my main way to track technical x product progress in AI (both open and closed, though i'd love to see a bit more open) His tech x product signal-to-noise ratio is *very high*

Google's made great strides, but people are still trying to make Gemini more Opus-like. What's the AI equivalent of the sincerest form of flattery? reddit.com/r/google_antig…

This might be one of the best written guides on how to build a deep research agent in 2026 I've come across (with a few finance specific things like fiscal year mismatches thrown in also). The real question for this year and beyond is whether these patterns can scale down in token cost and latency to be available for less pricey applications.

life is short build something that lasts

@TheGregYang Sorry to hear and hope you get to a full recovery!

I've been suffering from Lyme disease. I'm stepping back from xAI into an informal advisory role so I can go founder mode on my health, starting today. --- The symptoms started when I got sick (cold, flu, or COVID -- I'm not sure which) in early 2025. I distinctly felt less energetic, less creative, and less agentic even weeks after "recovery." After that, my condition ebbed and flowed, but the lows kept getting lower. Accidentally eating the wrong thing would make me extremely tired, taking days to recover. Working out would leave my whole body feeble for days. There was a week where I slept 12 hours a day and still couldn't recover. Lyme is famously hard to diagnose, but luckily I have an incredible doctor. He suspected these symptoms, far from being just in my head, indicated immune issues. Detective work over a few rounds of testing revealed I have Lyme disease. I was very surprised because Lyme is said to come from tick bites (where the bump looks like a target), but I don't ever remember having one. Likely I contracted Lyme a long time ago, but until I pushed myself hard building xAI and weakened my immune system, the symptoms weren't noticeable. --- Overall, I actually feel lucky to have discovered this early. Lyme is a serious disease that only gets harder to treat with age -- patients discovering it in their 50s or 60s have a much tougher time. Lyme can also be debilitating, leaving its victims bedridden, but luckily I'm still functional and can take care of myself day to day. So while some folks have said "you shouldn't have pushed yourself so hard," I'm glad I did. I found this issue early, and now I can fix it so I can push myself even harder when I rebound. --- Chronic Lyme is not well understood in the literature or by the public. For folks suffering from it, it can be a lonely fight. But I hope my story can make it just a little less lonely.

"The product sells itself" and "Sales is easier than building products" are my favourite laughable myths. PLG motion is excellent. Most companies should adopt it and do Growth initiatives. However: ▪️ Launching a product requires planning, good comms, excellent execution and a properly organised distribution. Posting on Linkedin & X with a nice video will not do anything. ▪️ If products sold themselves, Microsoft Hololens or Oracle Cloud would have been a hit already ▪️ Take Rabbit R1; it was an excellent campaign with lots of hype. Nothing is crazy good launching products, and Cursor & Lovable, the PLG kings. Salesforce is the best in B2B launch & hype. ▪️ Great sellers know what motivates their customers, how to reach their ICPs and when to be aggressive, be soft, go with the flow or just let it die. ▪️ Qualification is the toughest part of sales. You qualify every single time you outbound, negotiate, upsell, cross-sell or ask for feedback. ▪️ Complex industries are the toughest to crack; passion & drive gets you through the door but not a good deal.

@bfioca Am I the only one who turns memory off? :-) :-(

Note the Community Note, but still a testament to how far we've advanced in such a short amount of time. If anyone from OpenAI read this, please allow us to see Pro's thinking traces in shares: ChatGPT Session here: chatgpt.com/share/696ac45b…

Hence why decided to go all in on focusing on it