Michael Boeding

Software Engineer

[Michael Boeding @michaeljboeding]

If you’re having a bad day, here’s me getting beamed in the face.

[Michael Boeding @michaeljboeding]

@thsottiaux This pricing change is pretty rough. The old $200 Pro plan gave 20× usage. Now $200 gets 10×, while $500 gets 25×. Same price, half the usage, and you have to spend 2.5× more to get only 25% more usage than the old plan. The graph says it all.

[Michael Boeding @michaeljboeding]

@thsottiaux Is this for existing plans or only new subscriptions? Are we grandfathered in?

[ToughTail @ToughTail]

It’s finally here. 🐾

Introducing the ToughTail Standard Tracker — GPS tracking built to give your dog more freedom and you more peace of mind.

Because every adventure should lead back home.

Shop now:
https://t.co/uiPSHCmLsC

[Michael Boeding @michaeljboeding]

@thsottiaux Might be a nice idea to add a push notification when a reset has happened and when one is happening soon.

[Michael Boeding @michaeljboeding]

@babayagatwt It’s SSI

[Michael Boeding @michaeljboeding]

@thsottiaux It’s the auto review that’s using so much

[Michael Boeding @michaeljboeding]

@thsottiaux Latest update also doesn’t have the terminal option in the right panel

[Michael Boeding @michaeljboeding]

@thsottiaux Add read aloud feature to the codex output so we can listen to it. Like the regular ChatGPT has.

[Michael Boeding @michaeljboeding]

It would be nice to have an option to listen to the assistant’s most recent message, similar to the Read Aloud feature in ChatGPT. File contents should also be processed more intelligently for speech. For example, when reading a Markdown file aloud, it should interpret or skip formatting syntax instead of saying things like “dash, dash, dash.” Ideally, it would read the rendered content rather than the raw markup.

It would also be helpful to use Live Voice within an existing chat. That would let you talk naturally about the work already in progress, while keeping the full context of the conversation, and seamlessly switch between voice and text without starting a separate chat.

[Michael Boeding @michaeljboeding]

@jxnlco @jxnlco would it make sense to batch auto review items? Like have the model explain its next X steps and then review them all together so it’s 1 request instead of multiple?

[Michael Boeding @michaeljboeding]

@jxnlco

[Michael Boeding @michaeljboeding]

@thsottiaux @badlogicgames 20x

[Michael Boeding @michaeljboeding]

@thsottiaux @TokenGremlin Little confused the IOS app says there’s a Max but the codex Mac app doesn’t?

[GURGAVIN @gurgavin]

BOOM JUST AS I SAID LAST WEEK

SPACEX TO ACQUIRE CURSOR IN $60B ALL STOCK DEAL

CURSOR WILL BE A WHOLLY OWNED SPACEX SUBSIDIARY

THE DEAL IS EXPECTED TO CLOSE IN Q3 OF THIS YEAR

$SPCX

[Michael Boeding @michaeljboeding]

@claudeai One message used 99% of my usage lol? $100 dollar plan.

[GURGAVIN @gurgavin]

*META LAYOFFS SPARK PANIC AS STAFF RUSH TO GRAB FREE SNACKS, DRINKS & CHARGERS

*META PLANS TO CONDUCT MAY 20 LAYOFFS IN THREE BATCHES GLOBALLY

*META TELLS EMPLOYEES 'MANY LEADERS WILL ANNOUNCE ORG CHANGES' ALONG WITH LAYOFFS

$META

[Michael Boeding @michaeljboeding]

@thsottiaux When’s the rate limits reset?

[Michael Boeding @michaeljboeding]

@sama Intelligence > Speed

[Michael Boeding @michaeljboeding]

@claudeai Built this for Pro/Max plans per session since it doesn’t seem to be available for Max/Pro users https://t.co/PlS9CIDGz0

[Michael Boeding @michaeljboeding]

@nikitabier A beta build took over 8 days - usually takes a couple of hours for me….