ChatGPT's Trek Reboot Film Proposal

[Mr. Spook]

AI is a good tool if used properly, but too many people (read: the majority) use it for things they should be looking to professionals for or to doing themselves.

Look at all of the ad graphics that have the same look and stye of Chat GPT. That's how many graphic designers losing work? I know two personally who are.

Students using it for homework. Job applicants having AI scratch write their cover letters and resumes. You think employers can't tell the difference?

There are ads for AI writers: "type in your idea and it'll generate an entire novel!"

Plus the dangers of just this side of photo realistic recreations of real people for marketing and manipulation. We've come a quantum leap from CGI Peter Cushing in Rogue One. Deep Fake and AI are getting very close now.

Again, a great tool if you know how to use it and with good judgment. But it's too accessible right now. Without even touching the subject of environmental damage and wastes of resources, businesses are impacted by it when used in place of skilled individuals.

What should people use it for? If a person has zero skills at composing a professional reply to something and writes a draft, sure. Asking ChatGPT or something to "make this sound professional" is fine. Using AI to do line edits or whatever if you have nobody to help you. And, okay, you want to make some fun shit to share on your social. But I'd feel a little better about the fun part if it used fewer resources.

I don't condemn people using it but I hate the serious use of it in the place of using genuine talent.

 

[DonIago]

I recently edited a novel and found out by happenstance that when the author took it to a publisher, one of the reasons why they rejected it was because they thought it might have been AI-generated, and one of their reasons for being suspicious was that they didn't find (m)any typos.

When I heard about that last part I wanted to laugh and cry. "Of course there's few/no typos; I'm damn good at my job, TYVM!"

The one time I used AI and found it genuinely helpful was when my manager and her manager were both on vacation and I was asked to deal with a question involving calculating the correct amount of accrued interest on US Treasury Notes. I used Copilot, with reservations, to learn WTH I should be doing, because I was in about 5,000 feet over my head, and it did do a good job of explaining that (hopefully correctly!) to me.

 

[Tosk]

(hopefully correctly!)

Aye, there's the rub.

 

[DonIago]

Aye, there's the rub.

Well, I never got any response to my email, so I'm assuming no news was good news.

 

[Tuskin38]

Fuck AI

It's making people even more lazy and stupid.

Taking what it says at face value without checking its sources. It misinterprets things all the time.

Using it to do all their work without actually learning what it's doing, which is freaking dangerous.

 

[Mudd]

it can't be honest, brutally or otherwise, that is not within it's power. nor can it like doing it. it just strings together words it found that have been used in conjunction with the words you put in.

Sure, uh-huh, okay.

"Within its power" is an entirely inapt turn of phrase, BTW.

You probably want something like "Doesn't have that capability."

If you want to tack a "because" onto that and keep going, feel free.

But none of that parsing mitigates the fact that the words used in a prompt influence the output generated by the model.

 

[The Gatekeeper]

I wonder how many modern Trek episodes have a passing similarity to Trek novels ...

Data's resurrection in the novels "Litverse" continuity is very similar to what we end up seeing in Picard S3. Even down to the virtual realm which exists in Data's head.

 

[Ovation]

the fact that the words used in a prompt influence the output generated by the model.

Without a doubt. I test AI with prompts taken from my students’ work (I select material from students with various averages for grades). The work is the same topic for all students. The quality of writing is not. Feeding their statements of varying quality (but on the same topic) generates quite different results.

 

[publiusr]

How about Deep Fake adaptations of novels? ChatGPT might suggest related episodes.

 

[Deleted member 106352]

Current LLMs will write whatever you want them to write - utopia, dystopia, whatever makes you happy. They have a very large database to extract ideas from (science fiction books, films, series, fanfiction, etc).

Since they aren't the real artificial intelligence and don't really understand what they are writing about, they can write with a lot of inconsistency, they can become repetitive, even illogical. If people decided to use them to write scenarios, they would need to correct them and supervise them all the time.

I personally don't believe that the LLMs are the path to the true artificial intelligence, but they will get better and better at mimicking people, so at one point people might not even notice the difference. Everything online feeds them and improves them.

It's still not known whether they would improve to the point of inventing "original thoughts". Many argue that we also don't have many original thoughts and that most of "our" thoughts are the mixture of the previously learned ideas. However, we have one big advantage - each and every one of us is an individual being with the unique perception of the world molded by our experiences, our personality...

Current AI models have none of that. They don't exist as entities. That's why slightly different prompts can cause a completely different reaction, viewpoint, "personality"... They simply reshape into whatever you want them to be within the given guardrails. That's why we shouldn't take their "opinions" too seriously.

Even if this current technology never evolves into the real intelligence, I believe that it's here to stay. AI video generators are getting better and better and creative people started making interesting short films using them. Some people simply use them as a tool that helps them express the ideas they couldn't before due to the lack of skills, knowledge or money. Everything created with AI doesn't necessarily need to be "AI slop".

This means that we might get a lot of independent cheaply made "Star Trek" versions by fans who always wanted to create them, but didn't have the means and some might even become popular and gain attention. These new serials might become the continuation of the series and its savior.

I try to maintain the balanced view about this technology. I tried many AI models over the last year and a half for various tasks and I neither like the hype nor the doom and gloom. They can be helpful and useful, but also very incorrect and unreliable, prone to hallucinations.

 

[Mr. Spook]

Everything created with AI doesn't necessarily need to be "AI slop".

This.

And yes, the environmental impact needs to be considered. Over time, that will be ironed out. Or if they can't and it gets too bad, it will be behind large paywalls or have some kind of regulation. Right now, like the early days of the Internet and streaming, it's pretty free.

 

[Deleted member 106352]

This.

And yes, the environmental impact needs to be considered. Over time, that will be ironed out. Or if they can't and it gets too bad, it will be behind large paywalls or have some kind of regulation. Right now, like the early days of the Internet and streaming, it's pretty free.

I read a lot of different views on its impact on the environment. People disagree quite a lot about that.

It's difficult to predict the future. It will probably be implemented as a tool in many industries and AI chatbots will stay as digital assistants. Like any other technology, it will become a part of everyday life. It's still a novelty.