@shensi Cost savings are only half the story. The bigger win is choosing the right model for each task instead of throwing the most expensive model at everything. The workflow matters as much as the model.
@Prathkum Exactly. Coding is only one layer. Problem framing, systems thinking, debugging, trade-offs, and communicating with users become even more important when AI can generate code on demand.
@linear That matches what I’m seeing in practice: the biggest gain isn’t typing code faster, it’s shortening the loop from idea → first working version → feedback. The human bottleneck shifts to judgment, testing, and knowing what to ask.
Confession: AI writes a lot of my code at work now.
Less boilerplate, less repetitive work, more time for real problem-solving. I’m faster—and my boss has even praised the results.
AI didn’t replace my job. It replaced the boring parts.
Are you using AI at work yet?
AI isn’t magic. It’s a multiplier.
Give it a vague question and you get a vague answer. Give it context, constraints, and a clear goal—and it starts to feel like a very fast teammate.
The real skill isn’t using AI. It’s knowing what to ask.
What’s your best AI workflow?
AI will become as normal as using Google. We’re still early.
It can write, research, code, organize ideas, and help you get unstuck. It’s not perfect, but the time it saves is real.
The advantage will go to those who learn to use it well first.
What do you use AI for?
@thsottiaux could you please reset my Codex quota?To make this even funnier (and a bit embarrassing), I actually had to ask **Gemini** to write this message for me... simply because I’ve completely run out of Codex credits and couldn't even use it to type this out myself! 😅
Thanks a lot!