Sherlock

@ibuildthecloud @ThePrimeagen I think $100 is a little low as well, as much as I like it.

I really like @ThePrimeagen but this video is pretty far off base, but one thing people should accept is that the entry price for AI coding is going to most likely be about $100/mo for "professional level." Free or $20 is not in the cards. youtu.be/_Q-e_nczWqM?si…

@jxnlco everything is an ad smh lol

Take it from me, guys, you got to spend your arrogance while you're still young. You're gonna spend all your 20s trying to get there, and in your 30s you realize there is no there. You're going to realize that the skills that get you to the party aren't the skills that will get you to be happy with the rest of your life. You're going to go from quitting smoking to smoking again to quitting smoking in one day. You know what? You're just not smoking right now and that’s fine. The only constant will be time spend on the codex app will probs increase.

@normarayr @jayair Oh my. Thank you, I guffawed. Wouldn't have ever struck me (as someone with zero personal lines of credit).

@thesherlocker @jayair Line of credit. They want to give him a loan.

My dumbass trying to figure out how increasing lines of code at 8.99% annually would work

@yitong They killed my favorite font I loved instrument serif 🫪😭

behold the most cursed aesthetic of 2026: "Taste" set in Instrument serif over a midjourney generated impressionist painting not corporate memphis. more like noguchi table from temu vibes

@bruvimtired Git is good enough. Models are trained on it quite well, and you always havr @gitbutler if you want for for agents. (or jjj if you have that as your mental model) The human interface is still important here

this

@theo Most impressive thing is timing it within the bounds of the camera frame

Still got it

@yurinaessance @marcelkargul @figma It's okay bruh Project Risk and Ratios and stuff is equally as bad dw

@thesherlocker @marcelkargul @figma im getting cooked rn bro

terminal UI. made in @figma

These are so cool. We need a new kind of benchmark for visual IQ applied to these kinds of graphics. I’m still in complete disbelief that we can do this now.

this represents a bit of a fundamental misunderstanding of what's going on at AI companies. there's a bunch of chips companies use right now; almost no one's tied to a single architecture. but the most crucial spend of a company is training, not inference. Anthropic has a limited supply of Nvidia GPUs that they'd rather use for training (since Nvidia has a huge architectural advantage there) and wants to free up any of them currently occupied for inferencing. inferencing is a consumable product for each company, but training is their backbone. We obviously have exceptions like OpenAI using more NVidia GPUs with the GPT 5.5 series alongside traditional hardware and their other agreements, but that's also because they're not nearly as supply constrained on them.

@cneuralnetwork papercode pro with algo image gen /jk

@jxnlco did you try codex on all of them would it even run on something that's on Yosemite or older

@thesherlocker Really old MacBooks.

I’m limited by compute.

@CodingNoobie @CharlesLeFerg bro is lowkey one of the most positive accounts on the platform

@CharlesLeFerg Things will get better big man

I literally made this exact same post for 3,000 less likes but okay whatever

@reach_vb fr. chats need context and i'd slowly moved away from web stuff to coding agents over the last little while. lowkey super useful for exams

chats in codex are underrated, you can do som many one off, sheets, docs, search etc tasks without worrying about it polluting your project more people should use chats in codex!!

@SIGKITTEN @mweinbach no let it be; it's the curse of light mode

@mweinbach i gotta fix these blurry labels in light mode wtf

Damn @SIGKITTEN really did just add a fat cat to Litter

@larsencc air owns me?

If no one owns it, it owns you.

@badlogicgames okay now you're just reiterating your (great) talk at ai dot engineer lol it's such a useful thing to show everyone who talks to me with the "ai epiphany"

smilarly great signals: - managing agents is like managing a team of humans - but i have review agents - spec is all you need

i actually don't want this "but you don't review compiler output either" meme to die. it's the perfect signal for being immediately able to ignore someone in this space.

@ThePrimeagen @trashh_dev A friendly group githubs

@trashh_dev A friendly group gawaits!

universe is telling me something

@xlab_os @andrewqu Pretty cool! As a forgejo user myself, love the aesthetic, all the best!

@thesherlocker @andrewqu My license depends on the company size. I have Indie license with web use, but I am happy to make amendments if used incorrectly.

assuming reliability is solved, I tested many different git services for pure network-operation performance some shocking results for "headless providers" 👀 should I release the findings?