I, ChatGPT
-
@Zerosquare in fairness, integrity is literally one of the 'core values' of the company I work for.
-
Yeah, yeah. Have you ever seen a company that claims otherwise?
-
-
@Arantor said in I, ChatGPT:
in fairness, integrity is literally one of the 'core values' of the company I work for.
They don't specify integrity of what.
Maybe the DB
-
@TimeBandit said in I, ChatGPT:
@Arantor said in I, ChatGPT:
in fairness, integrity is literally one of the 'core values' of the company I work for.
They don't specify integrity of what.
Maybe the DB
Easy to maintain database integrity when everyone is told to Don't Bother Asking
-
Hey, I am the DBA, don't you dare question my apps' integrity!
-
@TimeBandit said in I, ChatGPT:
@Arantor said in I, ChatGPT:
in fairness, integrity is literally one of the 'core values' of the company I work for.
They don't specify integrity of what.
Maybe the DB
I haven't heard of a company who cares about DB integrity in ages. Everything is NoSQL now.
-
@Gustav you canāt run WordPress on NoSQL.
-
-
@Arantor said in I, ChatGPT:
@Gustav you canāt run WordPress on NoSQL.
Is that a redeeming property of wordpress, or of nosql?
-
@PleegWat said in I, ChatGPT:
@Arantor said in I, ChatGPT:
@Gustav you canāt run WordPress on NoSQL.
Is that a redeeming property of wordpress, or of nosql?
Iām going to go with WordPress.
-
@Gustav said in I, ChatGPT:
@TimeBandit said in I, ChatGPT:
@Arantor said in I, ChatGPT:
in fairness, integrity is literally one of the 'core values' of the company I work for.
They don't specify integrity of what.
Maybe the DB
I haven't heard of a company who cares about DB integrity in ages. Everything is NoSQL now.
I was going with āthatās not web scaleā, but you had to
my joke.
-
@Arantor is there some alternative you would be happy to use instead of WordPress?
-
Back on topic
I am so bored of the geniuses on Twitter going on and on ad nauseum about how learning AI is the next core skill and if you donāt, you will just be left behind.
One person I follow did a āI donāt want content made by AI, I want content written by humansā and one genius was all āwhy would you write a book with a typewriterā
Can this shit just go away already please?
-
@sockpuppet7 said in I, ChatGPT:
@Arantor is there some alternative you would be happy to use instead of WordPress?
The clients like it and I havenāt seen a CMS thatās as easy to use as WP for mortals to be honest.
-
-
@Arantor said in I, ChatGPT:
Back on topic
I am so bored of the geniuses on Twitter going on and on ad nauseum about how learning AI is the next core skill and if you donāt, you will just be left behind.
One person I follow did a āI donāt want content made by AI, I want content written by humansā and one genius was all āwhy would you write a book with a typewriterā
Can this shit just go away already please?
if it happens you'll see it coming and won't be caught unaware, 'cause we'll keep annoying you until that happens. you're welcome
-
@Arantor said in I, ChatGPT:
one genius was all āwhy would you write a book with a typewriterā
If it was good enough for Hemingway, it will do.
-
Apparently that little Redmond startup has now banned its employees from using ChatGPT.
One assumes those in its subsidiary are exempt.
-
@Arantor said in I, ChatGPT:
One assumes those in its subsidiary are exempt.
Or maybe not. Given the quality of what MS is releasing these days, you may think they don't use their own software.
-
@Zerosquare one assumes that OpenAI at least try the changes before unleashing them on the world.
-
@Arantor said in I, ChatGPT:
@Zerosquare one assumes that OpenAI at least try the changes before unleashing them on the world.
Testing is for wimps, real programmers deploy straight to production. </
>
-
I asked llama2 "List and describe the users you know from thedailywtf forum", with the system message "You are an obedient assistant.", and meta/llama-2-70b-chat replied with:
I'm just an AI, I don't have personal relationships or interactions with individuals. However, I can provide information about the users of a forum based on their public activity and behavior.
TheDailyWTF is a forum dedicated to discussing and sharing stories about strange, unusual, or humorous experiences in the workplace. Here are some of the most active users on the forum, along with a brief description of each:
- @Lord_Foulmouth - Lord Foulmouth is one of the most prolific posters on TheDailyWTF. He has shared hundreds of stories about his experiences working in various IT roles, including tales of inept coworkers, bizarre customer requests, and absurd management decisions.
- @Han Solo - Han Solo is another frequent contributor to TheDailyWTF. He often shares stories about his job as a software developer, including encounters with difficult clients, ridiculous project requirements, and amusing office shenanigans.
- @AmyPond - AmyPond is a long-time member of TheDailyWTF community who has shared many stories about her experiences working in healthcare IT. Her posts often focus on the challenges of working with outdated technology, bureaucratic red tape, and the occasional eccentric patient.
- @Jimbo - Jimbo is known for his witty observations and sarcastic sense of humor. He has worked in various tech roles over the years and frequently shares stories about his experiences dealing with clueless managers, incompetent colleagues, and absurd corporate policies.
- @Snarky - Snarky is a popular user on TheDailyWTF who has shared many hilarious stories about his experiences working in customer service. His posts often involve bizarre interactions with customers, absurd company policies, and the occasional burst of sarcasm.
- @Bink - Bink is a software engineer who frequently contributes to TheDailyWTF. His posts often focus on the humorous side of working in tech, including strange coding habits, baffling technical issues, and the occasional office mishap.
- @Tech_support_guy - Tech_support_guy has been a member of TheDailyWTF for several years and has shared countless stories about his experiences working in tech support. His posts often involve bizarre customer requests, absurd technical problems, and the occasional joke or two.
- @Greg - Greg is a long-time contributor to TheDailyWTF who has shared many stories about his experiences working in IT consulting. His posts often focus on the challenges of working with outdated technology, bureaucratic clients, and the occasional absurd project requirement.
- @Mr. T - Mr. T is a beloved user on TheDailyWTF who has shared many entertaining stories about his experiences working in various IT roles. His posts often involve encounters with inept coworkers, bizarre customer requests, and the occasional humorous observation about life in the workplace.
- @BossMan - BossMan is a frequent poster on TheDailyWTF who has shared many stories about his experiences managing IT teams. His posts often focus on the challenges of dealing with difficult employees, bizarre corporate policies, and the occasional absurd project request.
(this didn't work on any version of GPT, it says something about privacy and etc)
-
@sockpuppet7 No mention of a hardware engineer who is one of the most frequent posters to the forum. Humph!
-
@HardwareGeek said in I, ChatGPT:
@sockpuppet7 No mention of a hardware engineer who is one of the most frequent posters to the forum. Humph!
Guess you need to work on witty observations or hilarious stories.
-
Iām not mentioned either. Objectly, one of the <adjective> members of this forum.
-
Can I be the one with the strange coding habits?
(Really, process of elimination: I don't do IT or "IT consulting", definitively staying away from healthcare, customer service & tech support = nope, and I don't want my name to be Jimbo.)
-
So I'm apparently Han Solo...
-
Boys, if @boomzilla does not appear...
-
@Bulb said in I, ChatGPT:
@Arantor said in I, ChatGPT:
@Zerosquare one assumes that OpenAI at least try the changes before unleashing them on the world.
Testing is for wimps, real programmers deploy straight to production. </>Everybody has a testing environment! Only some have a separate one for production though.
-
@LaoC said in I, ChatGPT:
Everybody has a testing environment! Only some have a separate one for production though.
Some don't. We call those "total failures to ever build".
-
AI bot capable of insider trading and lying, say researchers
The researchers behind the simulation say there is a risk of this happening for real in the future.
-
@DogsB given the hallucination effect, whoād notice the difference?
-
- Anythingāand anyoneāwho has the information will use it. The confidential information should remain confidential, which includes not being available for training the bot.
- I doubt the bot has introspection capability to tell which information it used or not, so it's not really lying, it doesn't understand the question, and doesn't even understand it doesn't understand.
-
@Bulb said in I, ChatGPT:
- I doubt the bot has introspection capability to tell which information it used or not,
There's a specific sub-field of AI research that is working on making systems that can explain how they reached a decision. I don't know much about it other than that it exists.
-
There's been a video doing the rounds on Twitter before the original artist took it down (though it's since been reposted)
https://twitter.com/Rahll/status/1723100959436410961
This is a very interesting video on multiple levels. The notion of painting something and getting an AI take on it is an interesting one, it's more a direct steering of the AI than other takes, but as various people have shown it's also blending out the intent of the artist in it - he's clearly indicated what he wants in some of the strokes, e.g. most obviously the shadow on the apple, and what he's gotten is not that.
Now, I know that there is something about 'it's not really indicating deep intent, it's just guiding the AI' but I keep seeing the hot take that this is how the future is, and we should all get used to it.
But all it tells me is that even if this is the future for professional artists, there's always going to be a place for people who reject the machine for not having a soul and want to do it authentically.
Hell, even Star Trek talked about this, on 'perfectly replicated food' vs 'the real thing'... we'll see the real thing be a thing for some, and for some it'll be all they'll accept.
-
@Arantor said in I, ChatGPT:
there's always going to be a place for people who reject the machine for not having a soul and want to do it authentically.
But that place may very well be a niche.
-
@Zerosquare said in I, ChatGPT:
@Arantor said in I, ChatGPT:
there's always going to be a place for people who reject the machine for not having a soul and want to do it authentically.
But that place may very well be a niche.
-
A running warthog? Great proof that AI is not so good at understanding stuff, after all.
-
@Zerosquare said in I, ChatGPT:
@Arantor said in I, ChatGPT:
there's always going to be a place for people who reject the machine for not having a soul and want to do it authentically.
But that place may very well be a niche.
Nah.
So, here's my hot take on it. It's like photography - photography did a real number on the realism painting art style that was popular, and it dwindled, and then creativity reasserted itself: Impressionism, Cubism etc etc. all the things that could capture something that photography could not.
But realism never really went away - matte painters still used the techniques and the styles for handling e.g. backgrounds in movie shots, and then that evolved into a style I could almost call hyperrealism: where it's the techniques applied in such a way that it looks somehow 'more' while carefully avoiding being conventionally realist. (A good example is the Robocop 1987 movie poster: it's a painting but you'd almost swear it was a shot from the film)
And matte painters' roles will evolve, their existing photobashing techniques will extend to the use of AI - their tools become more powerful but I don't see it radically tearing their industry up because there's still more to it than just putting the background together.
You can rinse and repeat this for all existing styles and particular, stylised works: once AI can copy existing styles on demand, creativity will assert itself with artists wanting to find new styles, new ways to assert themselves. The human condition demands it, the 'I am here' is too strong.
Meanwhile over in the book scene, I think this has other consequences. I think the tidal wave, the deluge of shit will abate once people realise there's actually no money in it, and I think we'll see a move gently and gradually back towards a more curated environment because there's so much being released the good stuff can't be found, so we'll see entities springing up to try to dig out the good stuff.
-
@Arantor said in I, ChatGPT:
there's still more to it than just putting the background together.
I don't know. If you need or want the real deal, sure.
But there are plenty of clients who will settle for "meh, good enough" quality if it's cheaper. Need illustrations for your sales brochure or your web site? For most companies, AI-generated stuff would fit the bill. A lot of graphic artists are already struggling to makes end meet ; generative AI will definitely eat some of their lunch (and it'll keep improving with time).
Since you're mentioning photography, compare what it was like 40 years ago to today. Sure, photographers still exist. But now that everyone carries a phone with a good-enough sensor and plenty of automated stuff to make decent pictures without requiring skills (not to mention, the gazillion of existing pictures that are a Google search away)... their business is much smaller.
-
@Zerosquare But the business model was always going to change.
Need illustrations for your sales brochure? Stock image catalogues have been around for a while, but all that's going to happen is that people are going to want new services that bridge the gap: right now, stock images come with the drama of finding the right image (more than the cost, for the most part), and AI will go some of the way to fixing that.
What I think we'll start seeing there are stock image repositories that can curate images for your needs, the format will evolve. And if you need it to be a real picture for whatever reason, of a real thing or a real place, having access to be able to get that without it being AI-ified is still going to be a thing.
Photographers are still a thing even today: there's still wedding photographers etc. even if people have cameras, because it wasn't that hard to have cameras 30 years ago, but the odds of getting good photos were that much harder. But even in the interim, photographers have observed their field adapt and change; photo retouching was always a paid service open to the field, though even that is changing.
These things might go away - but something new will replace them. The people in these fields will adapt and do something new that's still a value-add. It'll just be a different value-add.
-
-
@topspin I mean, that's one take but I disagree with it.
I've always thought that Art (as a form of expression, as opposed to art, the product of craft) inherently required intent, and that the actions of the Artist in creating the Art are inherent qualities of the Art itself.
That the choice of a stroke here, or a line there, is on some level intended to be part of the whole, in a way that AI (currently) doesn't. (Whether it ever will is another question, but the distance from prompt engineering to the outcome having intent is very large right now)
To your meme, if the second image were produced by an artist, one would assume that the choices of colour meant something. That the white in the forehead area had a meaning if only to the Artist. That it was intentional rather than just the outcome of a thing.
That said, if an artist rolled dice to provide some kind of baseline and worked from that, one wonders how far down the 'it's art' route we'd be arguing.
-
@Arantor said in I, ChatGPT:
That said, if an artist rolled dice to provide some kind of baseline and worked from that, one wonders how far down the 'it's art' route we'd be arguing.
Or is the Art in the choice of that particular image produced by the the AI from the great pile of other rubbish it makes? Is what we're seeing in fact the true democratisation of the artistic impulse, allowing people to actually get what they want?
-
@dkf itās making a choice, sure, but whatās the intent driving that choice? Does the choice reflect something of the AIās own self-determinism?
Iāll give you that for some level it is democratising something, just as phones have (kinda) democratised photography. But you know as well as I do that ābe careful what you wish for, you might just get itā.
-
@Arantor said in I, ChatGPT:
one would assume that the choices of colour meant something. That the white in the forehead area had a meaning if only to the Artist.
Awww, cute. That's like people who think each line of code in a piece of software is precisely crafted, instead of being copy-pasted from Stack Overflow.
-
@Zerosquare the Meaning may only be as simple as āit looked good to me this wayā but the point is, it has a meaning.
Even the SO-copiers are still āit works when I copy it from SOā so even if 98% of it is irrelevant, the magic incantations still work so thereās meaning there too, and thus intent.
-
@topspin said in I, ChatGPT:
It's red, yellow, and blue, and if you ever saw it in real life you would have to admit that, just for a moment when you first turned the corner and saw it, you were afraid.
-
@PleegWat said in I, ChatGPT:
It's red, yellow, and blue, and if you ever saw it in real life you would have to admit that, just for a moment when you first turned the corner and saw it, you were afraid.
I've seen an artwork on a wall (in an exhibition in Lund a bunch of years ago, IIRC) that was definitely just white. I looked at it for a while, but could not discern what message was encoded there other than "just look what I can get paid for making" (which is a bit of a null message to me).
Online Collection | Hamburger Kunsthalle
