It's all about jokes, funny pics...

lost_in_c said:
I think base level ChatGPT, and probably many or all of the paid tiers, is there to harvest information from and not necessarily give it to you. The true power of those systems really comes out when they get focused in on an extremely specialized and specific field. Like producing proteins from scratch that actually fit the real world.
Click to expand...
How is that any different from say, MIT (ooh ooh, that's where Stallman is from!) ;)
 
Oh. Do other autogenerating software companies also harvest the internet?

Yeah I guess.

I'm just waiting for the first one to have the cojones to call themselves Skynet. Maybe Elon Musk will rise to the occasion.
 
I don't think anyone understood my MIT reference.

I'm talking about ChatGPT requiring a subscription to access specialized knowledge like law, medicine, and other knowledge repositories.

Before ChatGPT even existed, all that knowledge was in the hands of educational institutions like universities. Universities have libraries, and they have knowledgeable people conducting research and teaching - that is how knowledge is shared and maintained. All that effort - it's not for free! You kind of have to pay those professors. The very process of spreading the knowledge - some people are able to understand what they're shown and taught, and some are not. The ones who have PhDs are the ones who have that ability, the PhD is supposed to be proof of that. A PhD is also proof that one spent a lot of money to acquire all that knowledge - paid all that tuition to the university. And what did they get for that money? Access to information, and ability to make sense of that information.

Paying money (as in tuition) to acquire information - that is not a new idea. Paying a subscription to ChatGPT to access speciaized knowledge - that's merely automation of that same idea. If you don't have a PhD, how do you know where to look to find information you need?

MIT was merely an example of a hella prestigious university that charges very high tuition. Harvard and Oxford are other examples. Point being, it's universities like that, they are the ones collecting and maintaining repositories of knowledge. They are the glass towers that are expensive to access - but support specialists who can be consulted - for a fee. GhatGPT's subscription is merely an automation of that old idea.
 
astyle said:
I don't think anyone understood my MIT reference.

I'm talking about ChatGPT requiring a subscription to access specialized knowledge like law, medicine, and other knowledge repositories.

Before ChatGPT even existed, all that knowledge was in the hands of educational institutions like universities. Universities have libraries, and they have knowledgeable people conducting research and teaching - that is how knowledge is shared and maintained. All that effort - it's not for free! You kind of have to pay those professors. The very process of spreading the knowledge - some people are able to understand what they're shown and taught, and some are not. The ones who have PhDs are the ones who have that ability, the PhD is supposed to be proof of that. A PhD is also proof that one spent a lot of money to acquire all that knowledge - paid all that tuition to the university. And what did they get for that money? Access to information, and ability to make sense of that information.

Paying money (as in tuition) to acquire information - that is not a new idea. Paying a subscription to ChatGPT to access speciaized knowledge - that's merely automation of that same idea. If you don't have a PhD, how do you know where to look to find information you need?

MIT was merely an example of a hella prestigious university that charges very high tuition. Harvard and Oxford are other examples. Point being, it's universities like that, they are the ones collecting and maintaining repositories of knowledge. They are the glass towers that are expensive to access - but support specialists who can be consulted - for a fee. GhatGPT's subscription is merely an automation of that old idea.
Click to expand...
There is always Sci-Hub (and Sci-Net for newer stuff)
 
Reasonable answers, suitable for queries from public users, yes.

Professional-grade for being correct, useful and presented in a manner that is appropriate for the situation at hand - sorry, that much effort ain't free. No, it won't tell you how much for a kilogram of novocaine, that is for surgery departments only.
 
Interesting, I didn't know that. I figured that paying got you some higher level of answer, but didn't know what it would be. I don't use ChatGPT very much--not that I'm so brilliant, but I suppose it's hard to break a mindset of look it up yourself and only ask a question when you've gotten to your limit. In case that's not clear, got into the habit of instead of, I want to use named. Someone please tell me how, step by step, I'd first try to set it up, try to figure out why something wasn't working, and only, if I couldn't figure that out, ask for help. I hope that makes sense, I am sleepy today, and not feeling well, so if I'm incoherent, please forgive me.
 
scottro said:
...but I suppose it's hard to break a mindset of look it up yourself and only ask a question when you've gotten to your limit.
Click to expand...
If I can't find something after a few search re-words, it either can't be done or I need to get extra creative :p

Alain De Vos said:
I have ollama. Many A.I. models answers just for free. They give reasobable answers.
Click to expand...
My smart speaker can barely play music without doing something unexpected/random on free music streaming tiers, or the speaker gives me a spiel about how I'm not paying for a premium music plan, but here's some sort-of-related playlist instead. I tolerate it for free on a convenient speaker that sits in a room, but I like my PCs functioning consistently without trying to sell me something :p
 
scottro said:
Interesting, I didn't know that. I figured that paying got you some higher level of answer, but didn't know what it would be. I don't use ChatGPT very much--not that I'm so brilliant, but I suppose it's hard to break a mindset of look it up yourself and only ask a question when you've gotten to your limit. In case that's not clear, got into the habit of instead of, I want to use named. Someone please tell me how, step by step, I'd first try to set it up, try to figure out why something wasn't working, and only, if I couldn't figure that out, ask for help. I hope that makes sense, I am sleepy today, and not feeling well, so if I'm incoherent, please forgive me.
Click to expand...
Nah, you do make sense here. Let me try to respond.

When paying gets you a higher level answer - I personally imagine it to be like PubMed.gov, a database of medical knowledge available on the Internet. For free, you may be able to do a bit of search, and see that there are some academic papers written on a topic of your interest. If you pay, you get access to download the full-text PDFs of those papers. And if you have credentials of an actual doctor, you might get rather different results than a free public user, plus a bit of analysis - and you'd be expected to be able to judge for yourself if it's useful info or not. A real doctor would be more likely to have a paid subscription to PubMed than a public user, because that kind of investment makes sense for the doctor.

But if you want to learn how to work with named(8) - you're already a member of a community of experts. They will tell you to start by reading the manpage, and take off on tangents mentioned there. Setting it up? that's public information, and people can tell you if a given blog is a good start for instructions of what to do first, second, etc. If somebody is not an expert on computers, our answers won't make sense, they won't know where to start and how to troubleshoot.
 
astyle said:
I don't think anyone understood my MIT reference.

I'm talking about ChatGPT requiring a subscription to access specialized knowledge like law, medicine, and other knowledge repositories.

Before ChatGPT even existed, all that knowledge was in the hands of educational institutions like universities. Universities have libraries, and they have knowledgeable people conducting research and teaching - that is how knowledge is shared and maintained. All that effort - it's not for free! You kind of have to pay those professors. The very process of spreading the knowledge - some people are able to understand what they're shown and taught, and some are not. The ones who have PhDs are the ones who have that ability, the PhD is supposed to be proof of that. A PhD is also proof that one spent a lot of money to acquire all that knowledge - paid all that tuition to the university. And what did they get for that money? Access to information, and ability to make sense of that information.

Paying money (as in tuition) to acquire information - that is not a new idea. Paying a subscription to ChatGPT to access speciaized knowledge - that's merely automation of that same idea. If you don't have a PhD, how do you know where to look to find information you need?

MIT was merely an example of a hella prestigious university that charges very high tuition. Harvard and Oxford are other examples. Point being, it's universities like that, they are the ones collecting and maintaining repositories of knowledge. They are the glass towers that are expensive to access - but support specialists who can be consulted - for a fee. GhatGPT's subscription is merely an automation of that old idea.
Click to expand...
Oh. Then I think it was you that missed mine. I thought it was some triple-cross back reference or some deep thing, like MIT also charges you to harvest you, it was too complicated.

What I'm saying is that even if you pay for the paid services, which I have no doubt gives much finer answers than the public robot, they really offer it to you to harvest you, the paywall is a filtering mechanism, and the real money is elsewhere.

As for doctors and such, "journalists," etc, to what extent is it them using the software? In a couple of years it will dawn on companies that they can just run the machine themselves, and bye bye job.
 
lost_in_c said:
Oh. Then I think it was you that missed mine. I thought it was some triple-cross back reference or some deep thing, like MIT also charges you to harvest you, it was too complicated.

What I'm saying is that even if you pay for the paid services, which I have no doubt gives much finer answers than the public robot, they really offer it to you to harvest you, the paywall is a filtering mechanism, and the real money is elsewhere.

As for doctors and such, "journalists," etc, to what extent is it them using the software? In a couple of years it will dawn on companies that they can just run the machine themselves, and bye bye job.
Click to expand...
Educational institutions don't charge the same fees that commercial shops do. Money just flows differently there.

Don't confuse 2nd-rate content farms with databases that are hosted by NASA or John Hopkins or CDC.

Wired magazine does have some interesting tech-related content behind paywalls, but that doesn't compare to say, ACM databases. Wired costs about $10/mo to access. ACM membership runs about $300 per year and up. ACM databases are something that a professor or a senior engineer at HP or AMD would have a subscription to. Wired magazine - come on.

lost_in_c said:
They'll even be able to autogenerate a pretty girl to speak the words and "generate a precence."
Click to expand...
Done, those things appeared about 5 years ago.
 
It's ironic, isn't it? There used to be all this talk about robots replacing the working man. As it turns out, as we can see in labour farms like China, muscle power is more resource efficient than steel in more cases than not.

It's the management class that will no longer exist in 5 years.
 
Also, to speak of "second rate" when it comes to harvesting is to completely misunderstand autogenerated software programming.

There are, in fact, several tiers of harvesting. It's not that they farm the idiots on ChatGPT to feed the 7 figure a year engineers. They are just two distinct harvesting tiers.

More and more, the more advanced programs simply do not even care about mass data. They are refining and filtering their harvesting data more and more, that is the new goal. Predigestion.
 
astyle said:
Don't confuse 2nd-rate content farms with databases that are hosted by NASA or John Hopkins or CDC
Click to expand...
Sorry, were hosted by said actors.
We are going to have a new dark age where knowledge is going to survive the book burning and ignorance in some strange archives only to be re-discovered centuries later... In this case comming generations need to pry it out of the LLMs holding shards of it.
 
Crivens said:
Sorry, were hosted by said actors.
Click to expand...
'Were' seems about right in the sense that the hosting (of government-sponsored and scientific information, whose audience is the educated sector of humanity) has been mostly outsourced to large commercial hosts who get paid with grants from the government.

Once again, humanity is gonna look to the cloud for guidance... because the cloud is where 'intelligence' lives... hmmm. Not that different from when the tablets were made of clay instead of rare earths mined in hell. 🤣

That makes me a freakin' rain maker! 🤣
 
astyle said:
That makes me a freakin' rain maker! 🤣
Click to expand...
Well, you could, you know, sit down while "raining"? Just an idea?

What I meant was that you are at the mercy of some if that knowledge keeps being available. When it was books, you would need to go to every place and remove it. These days, some nutcase decides to remove it and away it is.
 
Crivens said:
Well, you could, you know, sit down while "raining"? Just an idea?

What I meant was that you are at the mercy of some if that knowledge keeps being available. When it was books, you would need to go to every place and remove it. These days, some nutcase decides to remove it and away it is.
Click to expand...
Yeah, with books, it's a bit easier if you've been keeping track of where every single copy went. On the Internet, that is next to impossible. Somebody's gonna download a PDF, make a torrent, and who knows how many people actually have the book. Yeah, tracing is not impossible with all the tools available, but you still have to know how to actually use them, and even then you'll make mistakes. The difficulties and obstacles got digitized together with the conveniences, y'know.

You realize that Google got started by a Russian and an American working together, and now an Indian guy is in charge over there...
 
astyle said:
You realize that Google got started by a Russian and an American working together, and now an Indian guy is in charge over there..
Click to expand...
Since this is the jokes thread - just check out who gave them their starting capital :D
 
You must log in or register to reply here.
Back
Top