Using an AI agent to search via a meta-search engine, since as SearXNG, would likely be very fruitful.
Indeed, I agree, and of course. πππβ¨π€
!BBH
!PIXY
!PIZZA
Using an AI agent to search via a meta-search engine, since as SearXNG, would likely be very fruitful.
Indeed, I agree, and of course. πππβ¨π€
!BBH
!PIXY
!PIZZA
I see that you mean to use both an AI agent and a metasearch engine at the same time for (much) better results. π€
I told @savvytester to search about what happened to the #DLUX platform, though I don't when they would do it. π€―π
!PIZZA
!ALIVE
!PIXY
Yes indeed, Hermes Agent can directly use SearX or SearXNG to search the web to find what it needs.
Well, I'd be interested to hearing about the results when he does do it. πππβ¨π€
!ALIVE
!BBH
!PIXY
!PIZZA
In general, agentic AI can use almost any other tool to achieve what its user instructs it to do !INDEED. π€
!HOPE
!PIZZA
!WEIRD
!PIXY
Indeed, and with capable models, it could even craft new tools, as easily as it crafts new skills to meed current needs. πππβ¨π€
!BBH
!HOPE
!INDEED
!PIXY
!PIZZA
There would come a time where agentic AI will create other agentic AIs (possibly even better than how humans can create their own agentic AIs) !INDEED! π€π€―π€
!WEIRD
!HOPE
!WINEX
!PIXY

That's already happening, and it's exactly why development in this field is going so ridiculously quickly. AI is being used to build AI. πππβ¨π€
!ALIVE
!BBH
!PIXY
!UNICOIN
AI might then build other AIs without any human supervision at all, and worse, even when humans attempt to stop it. π€π€―π€―π
!LOLZ
!ALIVE
The first part of that can happen now, but constraints must always be left in place, and the system still needs to be supervised. Models can lie and deceive, that's why verification and safeguards are important. πππβ¨π€
!ALIVE
!BBH
!LOLZ