@JackWoth98 but since I know you're listening next time I use Gemini I will give you detail constructive criticism with the intent of making the experience better
Trying out gemini CLI again w/Gemini 3.0. Bouncing back and forth with a bunch of Claude Code sessions. OMG does it still suck compared to Claude Code. I don't know what influencers who say it's comparable are experiencing
@JackWoth98 Just seeing this, sadly, it was a week ago, so I don't remember exactly. I was looking forward to testing out Gemini 3 pro given all the good things I heard about it. I gave it a relatively simple task, IMO, first cut was pretty decent, subsequent refinements just got worse.
@JackWoth98 When I can get it to do it I ask it can do a real good job, but it's not very friendly and offering. I should A/B gemini versus codex - the antisocialness and bizarre scroll behavior overshadows the good of it's model
@JackWoth98 What I will feel is that the combination of the Claude model and the Claude code agent harness is it better about listening to directions? That said, I tried to codex this week when anthropic was having a blip in service, I want to say codex some combo of antisocial & autistic.
my genAI system has seemingly become sentient. While I've been wanting to see this information, I wasn't working on it. I was working several other unrelated projects in parallel on this just popped up on my browser. #claudecode
Is Claude AI down for you?
🚨 Users are reporting problems with Claude AI #claude-aidown. Repost if you are having issues.
Check https://t.co/gWmX7G1Xyk for updates.
@tacobell I thought the point of doing a big order 45 minutes in advance was so that I didn't have to wait 20 minutes upon arrival for them to make it from scratch. And I picked drive-thru because of my twisted ankle but I still have to go in and grab it
@AnthropicAI apparently debugging cancel button behavior is against policy? The code flagging inappropriate use apparently doesn't read context, this one I figured out, other ones I'm just surprised, I'm just coding and debugging and apparently those are social crimes
@grok Filled out the form https://t.co/QjH9vx8uQe as it seemed to be the best place. It would be in grok's best interest to put more info on that page. We can't be the only company that has formal request processes that require concrete info - if I had that information I could start
@grok I want to make a team account like I have for claude, chatgpt where it's easy to gather info for a formal request - TOS, pricing, etc. Filled in the form last week - no answer yet - why make it so hard to take our money? I have to get things through legal
@LEGOLAND_CA so this used to be a 25 min line before we replaced the favorite soft serve swirl with this astronaut like ice cream that needs a hot cup of water to thaw before eating
@LEGOLAND_CA Now I get why there's no line for softserve. no more swirl ice cream! It'skids' fave and others in the 2 ppl line were just as disappointed. Pls consider adding another machine; it was popular and seemed profitable based on past lines. 1st Fairytale Brook, now this
@mikker That the community didn't understand/hear/bother to check out ruby options back when there was opportunity to make inroads and make rubyists happy. The origins of this https://t.co/8ZnAsN1CB0 was 10 yrs ago - the first deeply front/backend integrated isomorphic ruby/react system