Conversations with an AI

And who knows, if this has some success you may take it further and train an AI to be part of the fellowship. Wouldn't mind having a Data thirsty for knowledge here in the public forum (and I do wonder if there's a beta engineered by @Scottie and co in the private forum :-D awaiting launch).
 
Another proof of the “ohh so great“ capabilities of AI, this time Grok. It can’t even answer the most simple questions correctly:


Rant about AI summaries:

I saw various “great AI summaries“ from a number of tools and almost always I‘ve found it pretty bad, because it strongly tends to summarize things without any critical thinking as to the truthfulness/reliability of specific claims and the overall claims in the first place. I guess that shouldn't be surprising because no human is involved in such summaries. Another point for me that irks me quite a bit about many if not all the "great AI" summaries that tend to get used for evaluating stuff, is: An increasing number of people use it and share it instead of doing their own research and thinking, often with the result that even more BS gets into the minds of people. If you know the (quite humongous IMO) shortcomings of such AI summaries, I think they can be of some use, but I also think, increasingly, many people start to rely on such things with little to no critical evaluation of it, the result being, that things get distorted and more noise is introduced.

Also, it makes many people even more lazy in regards to putting some sweat and thinking into what they do. I just don't think that we are anywhere near to have real consciousness (let alone human consciousness) in any of those tools, which means that it is likely that many of those tools will continue to suck, unless you know exactly what you are doing when you use it.
 
I think there are 3 things that could make current AI a lot better and at the same time a lot less bad:

- A truly and completely independent tool that has almost zero propaganda/ideology/lies/agenda as background and where it is a fixed policy to never accept any of those things at the same time. Truth is the primary core rule. At the same time the tool should totally freely be able to access ALL data the internet has to offer (including any and all clandestine areas), in all languages/areas.

- A way that the just described totally free tool can easily and quickly access all ever written things in books, articles etc. that can still be read (in all languages) and include that into the database. And continually update the database with all new written things, like new books, articles etc.

- A way that this just described totally free tool can easily and quickly access all non written data, for example communication between people (like videos, interviews etc)
and include that into the database. And continue to do so.

Unfortunately, as far as I know, there is no such tool out there and probably never will be.
 
Last edited:
Back
Top Bottom