Use of AI in responding to posts

Using computers can be incredibly difficult with certain disabilities like the blind and vision impaired. A.I. in that instance assists people with those limitations, who might rely on screen readers and voice to text. So typing for them is slow

Hey @Dimity

I personally have mixed feelings, mostly negative, about AI altogether, and I feel that nowadays it is very common to question whether something was made or assisted by AI; regardless, these conversations are vital! 💚

Having an academic background myself, I remember how strict the rules were at university about the use of AI in our work. A massive fear of mine, one that I still carry, was having my work flagged for AI, simply because of the way that I write and think as a Neurodivergent individual. It is really frustrating that human processing can be mistaken for AI, but it doesn't help that we keep teaching AI to be more "human," if that's even actually possible.

While I acknowledge that some people use it as a tool to help them organise their thoughts or find the right words, I don't think AI could ever be truly capable of capturing the "humaness" as I call it, in the responses it gives. Of course, it can replicate the language of compassion because it has learned it from us humans who have taught it. But the AI doesn't actually "feel" anything, unlike us humans, who feel a range of emotions, and we put those emotions behind our words.

I will say that it is a valid fear to worry about AI-assisted peer/mental health support taking over services. I volunteered for a mental health organisation that removed its forums and replaced them with an AI service where people can ask questions. Losing that community was devastating, but it led me to be here with this community, so I have no regrets about it. Seeing real human volunteers be replaced by an AI program did teach me exactly what is at stake when we do prioritise automation over real human connection.

I hope what I have said makes sense to everyone 💚

@Dimity I first remember talking about AI in the early 1980s at uni ... so ... way back... he was young and enthusiastic, probably a decent chap ... I spent a lot of time with science nerds... married one... hadIT company in the 90s....and enjoyed some aspects of the growing tek field and the web, but also feeling out of my depth and paddling like mad to understand.. it wasnt really my field of interest... tho I had basic literacy in the field...I could appreciate substantial computing power could save effort and work ... hmmmm.

fast forward to now

nup, not mad about AI, especially not in sensitive things relating to human beings.

Yes I saw a little how LLM grew

they are just words

Human communication goes deep, can be preverbal, can be complex

I think @tyme once reassured me that using AI was against Sane practice and values.

I dont want to argue against those who find temporary relief from engaging with it, but I hope they dont overly push it as a panacea. It wont be. It took over 20 years for our government to regulate phones in the schoolroom. I had oodles and oodles of conversations with parents and teachers over the years, and I dont think big regulation is always the key, but we need to be wary.

The rise of various conditions associated with excess use of tek ... is only recent ... oooohhh... there can be lots of empricial research ... ooooohhhh .....but it takes TIME ... for real results to actually come in.

I have been told I sounded a bit machine like ... lol ... so much for all that ... so I can relate to the aspie aspect .... me trying to be clear and concise and factual ... or being human ... or relaxed ...

my personal story ... is developing a cervical disc bulge cos of overwork in my mid 20s, by a guy ... so called in love with the laws of physics, but unable to see when his dedicated, educated, human wife had been juggling gravity and children and paint brushes and tools, who he encouraged to be cheeky and treat me poorly ,,, cos I was not a math book. Note he was definite about it being math without the s .... as he was a yank .... from the den of the beginnings of the tek... I was only an Aussie... therefore backward.

Nup, tek and tek types can be callous, and boy did I think it weirdly suspicious when AI started boasting about being able to pass as human. So advocating deception as opposed to usefulness. Hmmm.

Mind you I appreciate and respect the attitude of scientific objectivity for the purposes of enhancing genuine growth in knowledge. But all humans have bias.

love your response

@MatchaToad

Thanks for your thoughts @MatchaToad @DogMan79 @Appleblossom . It's great seeing 'natural intelligence ' in action.

I also dabbled in IT in early years and witnessed a very early application of a precursor to AI for knowledge mining. I tried to replicate it and was shocked when the authors divulged they'd pre-filtered the repository with massive stoplists, which they shared with me. Needless to say their approach didn't work in any but their own specific domain.

I'd trust AI more if it were amenable to transparency and explication. Unfortunately some applications seem to magnify inherent biasses in training sets.

I appreciate there are some concerns, however I can attest to the fact it helps me express myself whilst keeping the facts true. There's nothing stopping people from manually typing out fabrications or copying someone elses writing.