Y’all, for 3 years it’s been the same seesaw:

Anthropic comes out with the best model on the market

Everyone rushes to Claude

Anthropic pushes usage limits usage too far

OpenAI comes out with a model that’s 90% of Claude’s model with fewer limits

Everyone runs to OpenAI

Repeat.

Claude, please shift the tasks you just added to Todoist over by one day.

Claude:

“Devised straightforward temporal adjustment strategy”

Whoa, easy there. We’re shifting dates, not adjusting our time machine.

“No em dashes” is not the right way to put it. “Structure your sentences so they don’t require em-dashes” yields a much better result. I find that the problem isn’t the em-dashes as much as the overly complex sentence structures.

Honestly Claude? Same.

Well, it took about three months, but I finally feel like I can walk away from my computer at 5 o’clock again. AI transforming my entire workflow was very disruptive to my Q1.

“Okay Claude, now rebuild the entire application to use GraphQL instead”. 😬

Boy, it won’t be long before we’re talking about the fact that nobody’s ever taking breaks because AI is always in their pocket. I can already see the backlash coming. Take breaks y’all.

Well, it’s got me pegged 😂

I know people are reporting that they’re getting better results with codex but man…the TUI for Claude is just SO MUCH BETTER.