Jordan Talks Everyday AI

don't miss your codex reset 🚨 made a lil open source mac app called Reset Watcher to track your banked codex resets and their expirations to give you insights if you should click the big reset button. on github have fun github.com/jordan-edai/co…

@angelbrodin @ajambrosino @gabrielchua i kinda wanna know the 1% of devs tokens that weren't in codex. that's the real story

@ajambrosino @gabrielchua can confirm it was hostile to me but i loved it anyway

and when this happened, it was with a version of the codex app that was actively hostile to teams that weren’t writing code. would show empty diffs, show script execution, etc

@jbentley 💀

@EverydayAI_ Anthropic lets Fable run product and Haiku run comms

Someone’s gotta study the public perception downfall of Anthropic. I’m so confused by this. Clearly multiple leaders OKed this.

Troubling. Anthropic now nudging for biometrics. 🔏 - On one hand, they say, 'Anthropic never receives or stores fingerprints, face data, or any other biometric information.' - On the other hand, Anthropic leaked Claude Code's full source itself, one missing config line published 513K lines to npm - A Discord group accessed their hyped 'bioweapon' Mythos, by simply guessing the URL from Anthropic's naming patterns - Anthropic sells safety as its edge, but both failures were basic hygiene, not advanced attacks - So..... are people gonna believe this and give the company that can't even keep its own secret sauce secret their biometrics?!

AI Just Went Multiplayer. Slack Is Where It All Comes Together x.com/i/broadcasts/1…

@ajambrosino monday desktop updates tibo mentioned dropping soon? (not like i'm refreshing twitter cuz i've got a training in the morning on it or anything. lolz)

team is cooking some great desktop improvements

@sama 🍿

come for the best model, stay because we don’t treat you with contempt

wait until people realize that GPT-5.6 sol is just a point upgrade and not an entire new model or even class 😆

Just wait until the conversation switches to ZDR It kinda matters, and valuemaxxing and ZDR are gonna own the narrative for the next few months. We'll see how Anthropic responds.

ChatGPT’s 5.6 Sol, Grok and Meta bounce back and OpenAI’s biggest week ever? And more AI News That Matters. x.com/i/broadcasts/1…

@cjzafir same. but yeah, that's why power users weren't hitting their 5 hour limits today. hahaha (i managed about ~3 billion this weekend and avoided it still)

@EverydayAI_ Its temporary. i wish it goes away permanently.

Pro Tip for Codex 20x pro plan: Use GPT 5.6 Sol Extra high + GPT 5.6 Luna Extra High and you won't any limit. Sol writes the plan, Luna executes, Sol reviews. Running this flow for last 48 hours and no 5 hour limit hit.

@SehjKashyap @gabrielchua Facts. I burned more tokens sometimes just to see which subs would pop up

@gabrielchua What happened to their fun names?

gpt-5.6-sol is now commanding a battalion of subagents

@gabrielchua Can we bring Jason and Dewey and sparticusmaximus907hgf pllleeeease 🙏

@dkundel @OpenAIDevs is the autocomplete agent dropdown new? don't remember seeing that before in the suggested replies.

@EverydayAI_ @OpenAIDevs That's just the title of a subagent task 🙈

I mighta missed this somehow... but did @OpenAIDevs just add Deep Research to Codex? If so bless the frick up making my runs now 🙏

@dkundel @OpenAIDevs i mean might as well make an agent shop then, right? lol (i've been tweaking my own version of deep research for codex prior to 5.6 but it seems like 5.6 ultra does the trick on its own. lol)

new number, who dis

@cjzafir saw it this morning cuz i have my limit %s permanently displayed in my menu bar and the 5hr one broke. hahaha

@EverydayAI_ lol i wish.

and just like that they're gone? maybe a test for tomorrow's update?