The AI assistant is backed by several different models which IIRC just call out to those providers (Op*nAI API, etc) and rack up tokens in the billing system:
You might be right that the AI cost is included when below the plan price — to that I have to say, give me a fucking cheaper plan that doesn’t implicitly include the cost of AI.
- 0 Posts
- 5 Comments
As far as I can tell, you have to completely disable all keyboard shortcuts or else when you press
Aanywhere that isn’t the search box you get dumped immediately into their AI assistant prompted with whatever you already had in the search bar.It didn’t track me more than a few pennies, but on principle the several times that happened made me angry.
Apparently some of the news views in search are also easy to dump you into AI land. There’s community CSS add on that hides all that stuff now, but I wish the company would let me just disable the AI traps.
Glitchvid@lemmy.worldto
Technology@lemmy.world•Is Kagi search engine worth it?English
462·14 hours agoI use Kagi because the truth is all other corporate alternatives at this point are unusable swill.
That said, I do not like the company and disagree with their choices in many aspects.
For one, while they don’t force you to use AI features, there isn’t a way to explicitly turn them off for your account, there always the opportunity to rack up token costs if you accidently hit one of the AI buttons.
They still don’t run their own index, instead complacent to just pay the other search providers. Additionally, if you’re trying to escape Google… Kagi runs on Google Cloud Services.
There’s more complaints, and I’m sure others will chime in, but that’s my take.
Glitchvid@lemmy.worldto
Technology@lemmy.world•US Government Urges Total Ban of Our Most Popular Wi-Fi RouterEnglish
3·4 days agoIt’s funny because I’ve had this one particular issue with two Asus routers that I manage for family…
They use this plunger power button design, you push the button in and it toggle locks in to place, the problem is that after a few years whatever mechanism retains the plunger fails and it always springs back and keep the device from staying on. So far the solution has been to cram a paper clip down the housing to hold it in. I just find comedy in having to apply that fix twice.
Do you have an example of this you can provide verbatim?
I’m just curious; I think the one application LLMs might actually be viable for is exactly this kind of connection finding in a large corpus, and since I’m doing lots of research, I might actually find personal utility.