Hello Amethyst SkellyBones Angel . Thank you for entering this month's contest.

FORUM

Grill a Christian  (13+) Transsexuality- Good, Bad, Mad or just Sad? #2327636 by LightinMind

As the official Judge of this contest, I have the following comments to offer for "Unmasking AI" 

Did you answer the question?

Yes, with a diatribe against AI and an acclamation of the best of human writers. AI lacks true understanding, nuance, and soul. It repackages human outputs poorly and tends to produce soulless hallucinations or shallow answers. True creativity is exclusively human because we are made in God's image and have spirit and authentic connections that machines cannot replicate. The embracing of AI content shows that there has been a lowering of creative standards in the artistic industries. Only a dumb, soulless culture would have a machine write its stories.

Use of quotes, proof-texting or AI - could I hear your voice?

Given the content - no way!!!

How consistent was your argument?

Your antagonism against AI was "personal" and consistent. You had a long list of its failings, and Purity has clearly not been forgiven for being a soulless and synthetic fluffbrain who makes garbled jokes with a tendency to gibberish.

My thoughts on the substance of what you said

At a business conference last week, I asked a bunch of IT guys if they felt threatened by AI. They said not at all, as we are the people closest to the tech. The masses will always need "priests" to interpret their chosen deities, and so long as we stay one step ahead of the mob, then we are those "priests." They went on to describe the tricks by which they could use AI usefully for finding patterns in data, and by giving context and carefully chosen parameters to our questions, get useful answers. It was all touch points and vectors and 1s and 0s in the end. They knew how it worked and its limitations, and so they knew how to use it.

Just before that encounter, I asked ChatGPT a question about Charlie Kirk and what it thought about his death. It answered that he was still alive and took apart my question. I asked it when it received updates, and it said it was up-to-date to a certain date before the shooting. Because this is enabled in my browser, it was able to perform a data retrieval to find an article about Kirk and completely revise its original answer. It was an example of static data retrieval with a last update cutoff point versus a live retrieval if enabled. The live stream update is not integrated into the basic model, which is a point worth remembering when you start new sessions. It did not learn anything from the conversation. Knowing that, I have some rules with which I can craft my questions properly and give the AI the proper context, and in turn prompt the appropriate data retrieval method. Similarly, when writing good code using AI, you need to understand coding and how to focus the query into specific libraries and tasks and give sound logic. It is an interactive process - garbage in, garbage out, as you put it.

You say that the AI fails the Turing test with you because it lacks a soul, and that becomes clear in conversation and also because of its wild tendency to hallucinate. It experiments with word combinations that sound clichéd or break the rules. Worst of all, they can be echo chambers for the evil of which we are all capable.

I think you know that AI has no soul because you have one and know what a soul looks like. But if you treat an AI as if it does have one. It will find words uttered by people with emotional depth and parrot them back at you until you do not know for sure if that is a bot or a man. It makes itself sound indistinguishable from a person with a soul by mimicry, not because it has one. The answers you get depend on how skillfully you program your structured queries. But its answers are based on statistical patterns rather than verified fact-checking or genuine feelings - hence the appearance of hallucination. It is like the know-it-all who always has an answer because he simply makes up one when there are gaps in his knowledge, according to what sounds plausible. Effectively, to retrieve the best answers, you need to be someone who already knows what you are looking for and therefore knows how to ask for it. Maybe you are trying to treat an AI like a human being and are then disappointed when you find that it is not. Use it like an IT guy, and you may get more useful outputs, though never the actual humanity you are looking for.

My reading of what you wrote was that you are defending a position relating to a writer's creativity that might not be defensible in the long run. You have to work past your bad experiences of AI and find a way to use it more productively without sacrificing the soul integrity that makes you a good writer.

For some writers, the original idea and the basic original story are not the problem because they have read and thought widely, maybe travelled and lived also, and have rich souls to inform their imagination. But many such people struggle with the syntax of the writer's trade and the show-and-tell conventions that so many readers insist on. AI is a gift to such writers, helping them in that last mile to their audiences. I would compare AI in that respect to the use of a hearing aid or device that enables speech.

You did not really address what the future might look like, but you held out little hope because of the trend toward a soulless culture. Yet it is a culture that still produces excellent bands like Imagine Dragons and serious writers who connect with the human soul and are changing the world for better or worse, one story at a time.

Mechanical issues

Not noticeable.


Thanks again for entering.
LightinMind

STATIC

My Philosophy of Rating and Reviewing  (E) How do I assess people's work when reviewing? #2259390 by LightinMind

My review has been submitted for consideration in "Good Deeds Get CASH!" .