Open AI/Chat GPT (1 Viewer)

CCFCSteve

Well-Known Member
Im no tech geek but the progress of this is pretty impressive/frightening




The worlds changing fast
 

Brighton Sky Blue

Well-Known Member
In my last year of teaching I used the more advanced versions to custom create worksheets and write reports. Could make a huge, huge difference to workload in a lot of jobs
 

ajsccfc

Well-Known Member
I had it write cover letters for me in recent months, quick proofread for clarity and off we go. Successful as well, but mostly because I'm a dazzler rather than the robot side.
 

messiahrobins

Well-Known Member
Not a huge fan currently as it doesn't save much time as what it does needs to be checked carefully. Colleagues use it for client care letters and the amount of times it makes mistakes is frightening.
However, eventually it will of course be nailed on and mean millions will be out of work as no longer needed. The idea of AI has in fact been around for decades. Although obviously fictitious programs, programs from the 80's such as Knight Rider and Star Trek the Next Generation (which had an android in it) were a nod to the advent of the idea of AI. In Japan many hotels now do have android type devices doing room service.
 

shmmeee

Well-Known Member
I already have clients being 'featured' in them. :cool:
I'm not convinced they'll last though, both from a user and legal perspective.

Didn’t they have this argument with snippets? I think the source links underneath will become important from both sides. I think their defence before was people clicking through.

Legally it’s a really interesting question. I think the theoretical economic incentives and huge amounts of cash involved might win the day for AI, there’s already countries that explicitly don’t apply copyright to training. Governments see it as a national security issue (personally I think LLMs are overhyped and it’ll come down to “does a lossy statistical compression of data breach the copyright of all the data compressed?”)
 

Potbellypig

Well-Known Member
It's just technology moving forwards and advancing into new depths. There are some great tools to help people in their jobs and hobbies, but nothing for people to be worried about.
 

Sick Boy

Super Moderator
It's just technology moving forwards and advancing into new depths. There are some great tools to help people in their jobs and hobbies, but nothing for people to be worried about.
Some of the Google AI answers are blatantly wrong; there's one about drinking urine to pass kidney stones faster.
 

Sick Boy

Super Moderator
Didn’t they have this argument with snippets? I think the source links underneath will become important from both sides. I think their defence before was people clicking through.

Legally it’s a really interesting question. I think the theoretical economic incentives and huge amounts of cash involved might win the day for AI, there’s already countries that explicitly don’t apply copyright to training. Governments see it as a national security issue (personally I think LLMs are overhyped and it’ll come down to “does a lossy statistical compression of data breach the copyright of all the data compressed?”)
Yeah, exactly - there was uproar about them at first too. They set the groundwork for the transition to AI a while ago with the aesthetics of the search results and even started highlighting text in the FS. The source is hidden by default in the AI answers unless you click on an arrow; I also have a client whose images are being used with no attribution.

My worry with it is if publishers have less incentives to create content, we'll end up with AI content based on AI content...though that is already happening to an extent, I guess.
 

wingy

Well-Known Member
It's just technology moving forwards and advancing into new depths. There are some great tools to help people in their jobs and hobbies, but nothing for people to be worried about.
Depths, don't bode well, heights suggests improvement?
 

Users who are viewing this thread

Top