Which AI Are You Actually Using, and Does It Matter?

There’s a thread doing the rounds comparing the major AI assistants, and it’s the usual mix of genuine insight and confident nonsense. But buried in there are a few observations that stuck with me.

Someone mentioned their mum now uses Gemini daily, gets answers in her own language, solves her own problems. Someone else’s mum has apparently made Claude her best friend. This is not the AI adoption story that gets written about in tech publications, but it might be the more important one. Quiet, practical, personal. Not productivity gains or enterprise integration. Just: my mum can get help now when she needs it.

That’s not nothing. That’s actually quite a lot.

The thread broadly sorts into camps. ChatGPT for research and coding, Gemini for everyday chat and multimodal stuff, Claude for conversation and nuance, Grok for… various things I won’t enumerate here but the thread was fairly candid about. There’s also a genuine debate about whether ChatGPT has gotten worse over time, with one person raising the theory that LLMs degrade when their training data starts including outputs from other LLMs. I don’t know if that’s been confirmed, but it would make a horrible kind of sense. A model trained partly on AI slop, producing slightly sloppier outputs, which then get scraped and fed back in. The quality doesn’t fall off a cliff; it just slowly softens.

I’ve been using a few of these tools seriously for about eighteen months now. Claude for writing tasks where I want a second voice that doesn’t just agree with everything. Gemini when I’m doing something visual or need to pull information from a document quickly. I’ve mostly stopped using ChatGPT for anything other than the occasional image edit, mostly out of habit rather than principle. The “talking to HR” description from the thread is unkind but not inaccurate. There’s a flatness to it now.

The thing I keep circling back to is the person who said Claude has the highest EQ, but then immediately flagged that normalising AI as a replacement for human connection might be “digging our own grave.” Both things are true and they don’t cancel each other out. If someone is isolated, and a well-designed AI gives them something that functions like companionship, that is probably better than nothing. And it is also a structural failure that someone got to that point. Holding both of those at once is uncomfortable, which is probably why most takes on this topic pick one and run with it.

The environmental cost of all of this sits in the back of my head every time I fire up a chat. I don’t have a clean answer to it. I use these tools because they’re useful, and I am aware that “useful to me” is doing a lot of heavy lifting in that sentence.

What I’m fairly sure of: the AI landscape is nowhere near settled. Whatever ranking feels true right now will probably look different in six months. The people who’ve planted flags loudly for one model or another are going to look a bit silly at some point. That includes me, probably.