Jump to content

Recommended Posts

Posted

Well I do have a friend who is a bit of a geek or 'fan' with those new gadget tech crap and who've briefly tested me with it with his talking "smart" phone (he should have known better) because I've evidently fooled IT for what it is ; a program. A sophisticated one but still that. That can seem impressive to who ever is by giving most acknowledge or generally accepted as being the truth and/or lies about science, culture and western politics...

Anyway I still hate the thing (and all such gadgets tools) for many "useless" applications (or future ones by replacing human critical (humor), wit (smart) and cynical (alive) work in the "communication" field for example) even though I know for fact that AI can have it's real purpose in accelerating or eliminating human error factors in some very tedious, complex and demanding resourceful scientific tasks for example that will benefit many.

So Dear Professor Loma today's class course proved once again a success very convincing of what such a program whose only at its early days will be capable to do. For the realistic conscious soul that I am it's just forwarding 'dooms day'. Knowing the worst of 'ChatGPT' is just around the corner. And I'm not talking about how many people's jobs are going down the drain in the years to come. That will be a social disaster on its own.

Nor am I talking about the 'political correctness' imposition in what ever form or the generalization of 'the single thought' watchful eye that such device will evidently bring by erasing moral intelligence and the fundamentals to replace it with modern liberal dogmas. Or the political propaganda it will force onto us all and especially onto those equipped with their private submissive devices in their pockets steady for the final consumerist Sacrifice.

The Merging ! One brain for all. Synthetic life in eternal happiness free of discomfort, conscience and privation. ChatGPT meets the MultiVerse. Every politician dream come true with all consuming and voting the same. Making it all not only easier and cleaner but also safer. The Dream come true. Technology finally prevailing over nature's approximative benefits and its unconditional unpleasant fatalities where all can be pre-programmed.

No am taking about more serious matter ; the 'nitty gritty' we're still argue about on here between how and what we love and don't love about the "scenes" or 'sceneries' (as in crime) for a more appropriate qualification if ever in plural, complex and most importantly in subjective forms. That with great-low coward caution 'ChatGPT' was programmed to "warn" us from. Reminding us that he's only that a non subjective AI incapable of emotions 👹

  • Up vote 1
Posted

I agree with most of what you say and concur that what AI in this context is today is but a shadow of what it will become - and I do find that scary!

Relative to my exchange with the chatbot, it is - as I said at the end of my post - very clear the AI is drawing from the populist aspects of what it finds on the web. The clever, the very clever, part is that it has the ability to present it in a cohesive, structured way. Sentence structure, grammar and logic can't be faulted, and that by itself is very impressive considering it takes but a few seconds to respond to questions.

There are obvious fact-based flaws, of course: "...dress in a distinctive Mod-influenced style...", for example, along with some other stuff. I've read worse in music mags and mainstream media that's been written by supposedly intelligent journalists (perhaps an oxymoron these days).

Copy and paste any sentence and use Google's Advanced Search to see where the bot got it from (on the basis it's copying from websites) and you won't find anything. The bot has very quickly amassed information on the subject, determined relevancy, and presented its findings. Again, that is incredibly clever!

 

  • Up vote 2
Posted (edited)
50 minutes ago, Amsterdam Russ said:

I agree with most of what you say and concur that what AI in this context is today is but a shadow of what it will become - and I do find that scary!

...

There are obvious fact-based flaws, of course: "...dress in a distinctive Mod-influenced style...", for example, along with some other stuff. I've read worse in music mags and mainstream media that's been written by supposedly intelligent journalists (perhaps an oxymoron these days). 👌

Copy and paste any sentence and use Google's Advanced Search to see where the bot got it from (on the basis it's copying from websites) and you won't find anything. The bot has very quickly amassed information on the subject, determined relevancy, and presented its findings. Again, that is incredibly clever!

Flaws are bound to be abundant willingly or not but stupidly always (like their "reliable" sources for the Mod-thing or the X-over def. also a joke) but ChatGPT developers will certainly re-inforce it's development in perversions for even more complex phrasings in 'acquired certainty' to fool most known as the 'crowd'.

Knowing like the 'human behaviorists' that these developers are that only few critics won't measure to the spread out lies established then as the new 'truth'. Plans that have always worked. Inducing fear (like a menace) to weaken one's brain capacity to evaluate reality better and develop anger to all now representing a 'treat'.

But now the fact that's an AI is behind it that is almost unidentifiable but maybe for an idiot... Maybe makes it scary. Seemingly human but not. We had the fake seemingly real imaging now comes the presentation with an argumentation. Already used by politicians while debating 'live' passing as know it all with the sources flaws...

Edited by Tlscapital
Posted

Most of those answers are very familiar, so I assume the wording is directly from internet text sources.

Simply typing the question into Google, or another search engine, would give similar results. I don't do it, but you can talk to Google and get audio results. Can you with this ChatGPT ?

Either way, at least with a search engine you can sort through the crap and concentrate on the real info. With ChatGPT it would appear the sources are not available so you're left with no options to evaluate. I would have expected to see phrases like "According to abc.com" or "xyz.com says that . . . " .

Can't see me ever using it.

 

 

 

Posted
4 hours ago, Mick Holdsworth said:

Most of those answers are very familiar, so I assume the wording is directly from internet text sources.

But that's the thing - as I mentioned earlier, they're not lifted from websites. I copied and pasted a few choice phrases to check. The responses are drawn from the AI's own database of stored knowledge, hence why there are no references to web sources: it's not quoting web sources. 

As far as speech is concerned, I believe that is possible. You speak to it and it speaks to you.

Early days in the life of ChatGPT. I dread to think what it'll be capable of in just a couple of years.

  • Up vote 2
Posted
5 hours ago, Mick Holdsworth said:

Most of those answers are very familiar, so I assume the wording is directly from internet text sources.

Simply typing the question into Google, or another search engine, would give similar results. I don't do it, but you can talk to Google and get audio results. Can you with this ChatGPT ?

Either way, at least with a search engine you can sort through the crap and concentrate on the real info. With ChatGPT it would appear the sources are not available so you're left with no options to evaluate. I would have expected to see phrases like "According to abc.com" or "xyz.com says that . . . " .

Can't see me ever using it.

 

 

 

Yes and no Mick, it actually uses a much more complex algorithm than this and evidently sources it all from the net but manages to avoid edgy, controversial or irrational versions and so manages to provide an approximative yet acceptable simple answer to any 'simple' questions one may ask.

Nothing relevant to anyone in the know of it if ever or whose deeper knowledge on the subject who might still object to this or that but to noobs it gives a first approach in understanding any matter without being that far from the truth. A vulgarized introduction like a teacher would give to its class. 

But most of all with phrasings and articulations of it's own in coherence attesting that it's not a mere 'copied and pasted' lines from various sources. Making it hard to tell for anyone at some point to tell if it's 'man or machine' made. Hence the issue for student essays done that way to tell if a 'cheat' or not.

So no 'copyright' payback here making it another problem for those who expected to cash-in on that. And it's only at it's initial phase... That alone in the A.I. world bluffed many. It's not another anecdotic 'software' nor a 'simple' development. It's considered a breakthrough in that field just as it's opening a new can of worms.

  • Up vote 2
Posted

Many thanks for explaining AI and ChatGPT in this context. Was thinking about doing something like this myself but struggled to find the right words (mind, english is not my first language).

I think the most important feature about this AI is that it doesn't just return any information available on the internet 1:1. ChatGPT is not a search engine (even though search engines are also AI). It reads information, analyzes it and draws it conclusions.

I use AI in my daily work. But only very basic AI's to detect clusters in big data. I find ChatGPT both fascinating and frightening.

 

  • Up vote 3
Posted
14 hours ago, Amsterdam Russ said:

But that's the thing - as I mentioned earlier, they're not lifted from websites. I copied and pasted a few choice phrases to check. The responses are drawn from the AI's own database of stored knowledge, hence why there are no references to web sources: it's not quoting web sources. 

As far as speech is concerned, I believe that is possible. You speak to it and it speaks to you.

Early days in the life of ChatGPT. I dread to think what it'll be capable of in just a couple of years.

 

12 hours ago, Tlscapital said:

Yes and no Mick, it actually uses a much more complex algorithm than this and evidently sources it all from the net but manages to avoid edgy, controversial or irrational versions and so manages to provide an approximative yet acceptable simple answer to any 'simple' questions one may ask.

Nothing relevant to anyone in the know of it if ever or whose deeper knowledge on the subject who might still object to this or that but to noobs it gives a first approach in understanding any matter without being that far from the truth. A vulgarized introduction like a teacher would give to its class. 

But most of all with phrasings and articulations of it's own in coherence attesting that it's not a mere 'copied and pasted' lines from various sources. Making it hard to tell for anyone at some point to tell if it's 'man or machine' made. Hence the issue for student essays done that way to tell if a 'cheat' or not.

So no 'copyright' payback here making it another problem for those who expected to cash-in on that. And it's only at it's initial phase... That alone in the A.I. world bluffed many. It's not another anecdotic 'software' nor a 'simple' development. It's considered a breakthrough in that field just as it's opening a new can of worms.

Okay I see now it is far more sophisticated and complex than I understood (or ever hope to understand).

As a layman, I suppose the point I was trying to make was that reading some of the AI responses on a subject we are all aware of, I can't help but feel it is just saying the same things we've heard before, but in a slightly different way.

I see from your explanation that it makes sense to answer in a way that is useful to the reader, without giving an opinion.

Cheers for that, both of you.

PS

I'm sure you're aware that ChatGPT now has a new version: GPT-4. I only saw it on the news feed last night.

 

 

  • Up vote 1
Posted (edited)
On 14/03/2023 at 19:00, Amsterdam Russ said:

 

On 14/03/2023 at 19:00, Amsterdam Russ said:

You
I want to be a Northern Soul dj. What 10 Northern Soul records should I play?

1. Tainted Love - Gloria Jones
2. Do I Love You (Indeed I Do) - Frank Wilson
3. Out On The Floor - Dobie Gray
4. The Night - Frankie Valli & The Four Seasons
5. At The Top Of The Stairs - The Formations
6. I'm On My Way - Dean Parrish
7. What - Judy Street
8. Time Will Pass You By - Tobi Legend
9. Landslide - Tony Clarke
10. You Don't Love Me - Epitome Of Sound

 

 

Brilliant!🤣

Now i know where some DJs get their playlist ideas.......

Edited by Source
tidy up the quote
  • Up vote 3

Get involved with Soul Source

Add your comments now

Join Soul Source

A free & easy soul music affair!

Join Soul Source now!

Log in to Soul Source

Jump right back in!

Log in now!


×
×
  • Create New...