🛜 haze

I fell lucky Gemni so far is not flooded by AI slop. The entire surface web is now that. And really low effort ones. Hope this continues.. I hope at no point there'll be so much people on Gemini that there's bot infestation

1 week ago · 👍 userfxnet, cobradile94, a1, eph

Actions

👋 Join Station

5 Replies

🌲 half_elf_monk

I guess we can say that the protocol doesn't seem to have a lot of *slop* ... whoever prompted the LLMs (or put together the agents) is doing a reasonably decent job. · 4 days ago

🌇 eph

@half_elf_monk very true. I wonder the same thing, actually; how many capsules are LLMs and how many are real people. It is comforting to know that most of what’s available in Geminispace is probably not thrown together by a large set of Markov chains ⛓️ · 5 days ago

🌲 half_elf_monk

+1 to what @eph said. There's not much reason to think that there aren't already a ton of chatbots/llm-capsules on this platform. Obscurity, maybe, and the good character of the people who DO find this protocol. But it'd be trivially easy to set something like that up, since this is mostly text-based. · 6 days ago

🌇 eph

Gemini would be easier to subvert than the Web with AI slop, but fortunately the Goog AI offering is running SEO interference, keeping us relatively safe (maybe). · 6 days ago

👽 lars_the_bear

I don't think the problem is with the web as such: there's no AI-generated content on my website. I don't see much on the other websites I tend to look at.

Moreover, I don't think it's helpful to divide content into "good" and "AI" -- there's a lot of rubbish written by people, too. Whether you use Gemini (etc) or the mainstream web, you just have to be discerning. · 1 week ago