bdlowery

yes, it's 2026, people use ai.

I don't want it hand coded, but you could put in some effort in and write more than 1 prompt lol.

coming from someone who is the farthest from an ai hater, these zero effort cookie cutter one shot claude sites are just bleh. Why would I bother using the site if you put in 0 effort

> This bench famously detected a degradation of Opus 4.6 which Anthropic later blogged about.

this bench was just released, it couldn't have detected opus 4.6 degradation.

You can tell what type of framing the article will be about AI by what social media they choose to use and link in the footer of their website. It never fails.

The fact that gemini 3.8 flash is so high up there just tells you this is an awful benchmark.

Try and use gemini 3.8 yourself for any real world work and you'll see it's terrible. It'll just go in circles reading the same file 20 times for no reason making hundreds of tool calls for a simple change.

EDIT: I was using gemini cli... it's not a harness issue lol

benchmaxxed model

Good.

> I don't want a full screen pop up telling me to sign in or a persistent bottom bar telling me to open the tweet in the app or a reply section mostly hidden unless I log in.

make an x account then?