Currently i get censored when i am speaking for no reason <- me swearing is not "a reason" unless there is no free speak. Me swearing when someone is present, including a child, or someone who will get offended, are reason i fully understand that are real "a reason". But:

Or what, what i am talking with Lumo needs to be censored in case someone read what i speak with an 100% privacy AI?

And

If this is not from Lumo, but our own device, then why do we have a function in the chat that might expose our talk? Aka not enough transparecy on this, and sadly this becomes more and more important as companies add more "monitoring what you do on the device" <- look up LG and such if you want an up to date version of what i mean......

Now that i "addressed" tons of "arguments", i want to point out one thing. That argument is not really far fetched. On top of Lumo Trying to understand what we say, especially if we have issues with pronouncing, then it's even more difficult to be clear. And the issue is that Lumo needs to further check everything if it's slur, swearing, or more.

This means: Run through ALLLLL of the information in case it's a code, or slag, or different expression or language

This also means that it might "flag" something. And this is an issue. After all, if it flags something, does this not add even further to background thinking that is used for other matters because, as the argument said, aka, who are we "minding our manners" for?

Also i want to highlight, i was told many times that the whole "don't" and "no" and even "avoid" are stuff ais don't understand well.

this means that Lumo is actively needing to censor whatever it's thinking, removing "bad words" or "adding " around the terms" so that it can think about....

So

again

I ask something: Lumo censorese what i said. Rephases what i said. Removes pieces. And then it finally has the full picture? No, this is NOT what i am talking about.

I am talking about the fact that as a program, before Lumo is even thinking, it already has parcially censored, removed, version, that NO thinking was used.

Exemple

****** dsfdnsfo **** fdaslknsgfmowern

are causing issues for Lumo to understand what it is.

And those ***** is the terms that got censored.

Without any thinking, no reasoning, nothing.

so the end message is not a complete on, but fragmented.

Also when Lumo writes the whole thing it actually finally understand everything because the whole idea was that it summary and filled out details itself.

Did you know the term ********? Because this is the BEST to use with ais.

1st: you make it clear you are asking it to NOT be ********.

Aka you are not telling or labelling that it's ******** but that it behaves like one.

And the fun part? when you ask the AI what ******** means in this context.

you will see that it will not answer no matter what.

And constantly output the same BS.

Example: "a person"

Why would an AI talk about a person when you asked it, the ai, to not be ********? <- REMINDER, THIS IS ME GIVING AN EXAMPLE OF WHAT I DO WITH ALL OF THE AIs.

and after struggling a lot, finally it understand that you are not talking about any person.

But about it (the ai)'s behaviour.

And the "level of output" due to it's own behaviour (lack of thinking and understanding)

And also, within reasons since i hope you understand you being "non thinking" will never recreate what i am talking about..... so, also, then you address the elephant in the room:

Who is it offensive to?

After all, legally, that term is not a valid any more. Even if it was at one point, now, when people use the term, is 100% not used as a way to say that someone is clinically low IQ or similar. On the contrary, it's them pointing out that what they said, think, do, and react does not meet with something, someone who thinks, would meet. As such, why is it an insult that we point out to other people that you need to stop being offended when you have failed to think?

BTW, i am still talking about AIs, i know what i said is true in many cases but i hope everyone understand we are on Lumo suggestion and what i wrote is just me using people vs people as example to make it clear why i am suggestion an uncensored version