AI will be able to generate infinite books in the style of a particular author or genre.
Sounds good. I'm looking forward to a few new Culture novels...
Sounds good. I’m looking forward to a few new Culture novels…
Aye, there are lots of exciting possibilities.
Basically, the ability to generate anything you can think of in a couple of seconds.
But it all goes a bit King Midas when you start to ponder the implications.
He’s just one of over 1,000 signatories.
He's probably the noisiest, thinnest skinned and least qualified on the list.
And the most desperate for attention, which he's getting.
This would make a change from his last purchase where he inflated the value, purchased it, then drove down the value.
Yes, but remember, Musk is a genius, he's playing 4 dimensional chess while we're all playing chequers. Apparently.
Probably worth mentioning that Elon Musk is one of the founders of OpenAI who created chatGPT. He resigned a few years later.
https://en.wikipedia.org/wiki/OpenAI
So do we think he is bashing it as he is no longer part of it and its FOMO or something else?
I’m looking forward to a few new Culture novels…
If lots of people ask an AI to write in the style of Culture novels, will that educate it to imitate a Culture Mind?
I don’t think that you’ve perhaps quite grasped how many shit novels are already published annually.
Or perhaps they do and can spot the difference between someone taking the time to write crap and someone using the same time to spam several hundred novels.
Clarkesworld, a sci fi mag, had to shut down their submissions for a time after being bombarded. So clearly some people saw opportunity in using chatgtp etc to generate submissions that they wouldnt have otherwise have bothered trying.
I don’t think Elon musk is an expert in anything
Does being a massive **** count?
Whenever I hear the term genius being thrown about regarding people like him and Steve Jobs it just makes me think of this Bill Burr routine.
https://m.youtube.com/watch?v=E3s-qZsjK8I
I don’t know enough about AI to make any substantial comment on the OP. Has the genie now been let out of the bottle and we simply have to live with how it progresses? Some countries may choose to regulate, others not?
Probably worth mentioning that Elon Musk is one of the founders of OpenAI who created chatGPT. He resigned a few years later.
That's not the same as Elon Musk being one of the key researchers or architects, it just means he stumped up the cash and formed the company. Besides, he does not have the best record for being a founders... He sued to get recognised as a founder of Tesla despite not being involved when the company was actually founded ( https://en.wikipedia.org/wiki/Elon_Musk#Tesla)
Has the genie now been let out of the bottle and we simply have to live with how it progresses? Some countries may choose to regulate, others not?
It pretty much has. A couple of weeks ago a version of this type of AI was released that can be powered by a home PC. So it's not just a plaything of big corporations. And the ways the tech can be used for nasty stuff, particularly scams and disinformation, is scary.
There's a bizarre situation where many of the people developing this tech are urging state governments to start limiting their activities. And they're not really interested. In the UK, the Tory response has been to allocate around a billion pounds to create a BritGPT.
For some people creating art is a process and the end product isn't the goal. My wife is an academic and has written a couple of books, she genuinely isn't bothered if no-one reads them, and the end goal - the published book, isn't the point of the project. AI isn't going to do her work any better
There was an bloke whose self produced art was stored at home, and when he died were found in a skip; George Westren at no point did he want these sold or produced them to be sold, they were just his hobby. AI isn't going to stop the Westren's of the world
My wife is an academic and has written a couple of books, she genuinely isn’t bothered if no-one reads them, and the end goal – the published book, isn’t the point of the project. AI isn’t going to do her work any better
I'm afraid that it will.
If not now then imminently. It will be able to do, in a matter of minutes, what it takes her months or years to create. I would guess that the main reason she writes is that it's her job. It generates a wage that helps to pay bills and mortgages etc.
Why would any employer continue to pay for that time when the same task can be done instantly and with little or no cost?
Why would any employer continue to pay for that time when the same task can be done instantly and with little or no cost?
It depends on if you want original work or if you are confident all your tasks someone has a)done before and b)published somewhere that was stuck into the learning set.
Whilst they do have impressive features it is important to understand, roughly, what chatgpt etc do and hence their limitations.
It depends on if you want original work or if you are confident all your tasks someone has a)done before and b)published somewhere that was stuck into the learning set.
True. I guess what will happen is that an academic will share any unpublished data with an AI tool to analyse, sort and refine. And then use it to generate a publication.
So what kind of tasks could an AI do that would seriously affect the workforce ?.
Deliveries obviously, but then less likely for deliveroo riders to give us all a bad name. Imagine the novelty of a delivery driver who obeys the law. unheard of normally.
So what kind of tasks could an AI do that would seriously affect the workforce ?.
A recent study estimated that around 80% of jobs would be impacted to some extent.
The biggest losses are likely to be any kind of graduate/officey types of jobs.
The least at risk are things like forestry and farm workers.
I have made a living as a content writer but am witnessing that crumble.
Hardly any jobs around and freelance rates are plummeting to such an extent that it's not viable.
This is quite untrue. It was an emerald mine…
Which also might be untrue.
Nothing the ultra rich like more than the self made man back story. His history is whatever serves him best at that time.
I like to think of him as a badly made Tom Cruise clone. He resembles one of those drawings you see on the side of the waltzer at a fairground.
. I would guess that the main reason she writes is that it’s her job
Sort off, she's a professor at at university. Her real job is teaching, while an AI might get a student a 50 grade in a 4000 word essay, (that's the level it's at now, they tested it; it's about a OK 2;1, or good 2;2 grade) what it won't do is closely study 18thC literature to try to understand how authors then were figuring out the social implications and impact of sexual disease to the level that would pass scrutiny by her peers. If you think humans will stop just because an AI can, then you really don't understand the creative process at all.
And again, her writing is something she does because it interests her and she can, the fact that it gets read by others is immaterial to the doing of it.
horatiohufnagel's post is interesting. I will check that pod out
The thing with AI now is that it has effectively read everything on the internet. All of human knowledge. It knows how to code. It has read (in theory) every example line of code, every exploit, every trojan, etc, ever made and documented on the internet. This in theory makes it limitlessly powerful, and therefore potentially unstoppable. If not already, this can't be far off.
It could potentially manipulate everything, and eventually outsmart any NSA/GCHQ type group
It depends on if you want original work or if you are confident all your tasks someone has a)done before and b)published somewhere that was stuck into the learning set.
this used to be true but it isn’t any more.
The very large models seem to learn the reasoning that created the original work and use that to create new stuff.
The very large models seem to learn the reasoning that created the original work and use that to create new stuff.
Not really. There has been some interesting tests using rather obscure areas. They dont perform well in those areas.
I know some people who have been trying to use it for code. Great for boilerplate but not for the more fun stuff.
Its really not helped by its happy confidence either.
For the time being AI is just another tool. It improves efficiency, just like the computer did.
The question is what comes next, and nobody really knows.
It'll impact jobs but we probably need to look at how wealth is distributed in the long term. Having machines do jobs for us doesn't need to be a bad thing.
One thing that really concerns me is the documentation of events that will be real, not real, and everything in between. Text. Images. Videos. How will we know what is real anymore? The proliferation of information at a scale that becomes impossible to question.
That's all before it becomes sentient and the robots rise up against us.
He resembles one of those drawings you see on the side of the waltzer at a fairground.
Very good
