Catch up on stories from the past week (and beyond) at the Slashdot story archive
Nickname:
Password:
Nickname:
Password:
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
And companies using AI instead of Humans in a creative space are rapidly going to find the results unimpressive.
It’s like we noticed with car manufacturing, the very-expensive handmade cars, last a lot longer. Same with consumer electronics. As soon as things stopped being hand-made, the quality went down.
Seems the order of operations is :
1) Do everything manually
2) Outsource labor to a cheaper place to do the exact same thing, except lower quality
3) Automate it once the unit-to-unit variation is within an acceptable-to-the-customer threshold
Like take a cardboard box as an example. Boxes have been around for a long time, it’s down to a science. Robots can do it. There is no need to hire humans to make boxes. Yet, when we apply this same logic to consumer electronics, we end up with stuff that could never be done by hand (eg chip manufacturing) and the human element moves entirely to the QA and Maintenance.
And this is where “AI in a creative field” falls on it’s face. These companies are not going to pay people to QA the output. We’ve already seen they do not. The only way to hold “AI in a creative field” in check is holding companies liable for producing factually incorrect information and not being permitted to “blame it on the AI”
The AI is not a “get-out-of-responsibility-free” card. If you replace your writers and artists with AI, and they plagiarize another companies work. That is your problem for not having checked it.
Unironically the opposite of reality. Cars became more reliable as humans were taken out of the manufacturing chain. No human can compete with a machine in accuracy of pain application. No human can compete with a machine in accurate placing of components onto a circuit board. Etc. Machine can do the exact same task in exact same heavily optimized pattern forever. Even the best, most motivated human has bad days. There’s a reason why concept of “thing made on Monday” isn’t really a thing any more like it us
” No human can compete with a machine in accuracy of pain application.”
“You had my curiosity, but now you have my attention.”
It’s missing a t 😀
Rather obviously, but hey, autocorrect didn’t correct that one!
While true, a huge chunk of work in graphics design isn’t particularly “creative”, and copywriting is really not.
They want something vaguely distinct and inoffensive for various visual impact, but with no ‘art’ behind it. The marketing drivel they want from copywriters is soulless uninformative garbage.
So no, not going to do well to make engaging fiction, moving works of art, and such. But it will happily churn out soulless corporate crap that those companies love.
1) Copyright is absurdly long, mainly for the benefit of corporations. If corporations benefit from shorter copyright terms, the terms will shorten.
2) I prefer the stuff written before 1950 already, and often prefer the stuff that’s already out of copyright.
That said, the authors and artists themselves were trained on copyrighted materials. Tell me how this is different.
P.S.: I’m not saying this isn’t a socially disruptive process. It is. It’s just that your argument isn’t based on any valid principle t
Copyright being long or short isn’t the point, the point is that AI simply gives back stuff that has been written by people.
Not generative AI, but regurgitative AI.
Eventually, when everything has been regurgitated, you’ll still need people to write something new.
You are assuming that statistical procedures produce meaning beyond what they get out of a data set. This isn’t true, and shows you simply don’t understand very well what is “AI”.

No, chatgpt isnt taking jobs from “the cream of the crop”. AI is taking jobs from the people who used to do the bottom-of-the-barrel, low-paid freelancing work. Need 3 unique general-info paragraphs about otters? You used to pay a freelancing english major to write it. Now, chatgpt can remix the 250,000 pre-existing otter essays on the web into something that’s just different enough to not violate copyright law. Same goes for graphic artists. Need a graphic of a soup can? Some generative AI can now do it. But this isnt taking work away from the best. Not even close.
I think you are correct. My experience with AI is that it can do a somewhat poor job.
There is a crop of AI generated videos on science of technology on Youtube. Overuse of Hypberbole, and not all correct. It really uses a lot of the “Scientists are stunned!” bs and you watch it and it’s like something about voyager that’s been known for years, and isn’t stunning at all, just reporting what the Voyagers have found. A history major could do almost as well.

No, chatgpt isnt taking jobs from “the cream of the crop”. AI is taking jobs from the people who used to do the bottom-of-the-barrel, low-paid freelancing work. Need 3 unique general-info paragraphs about otters? You used to pay a freelancing english major to write it. Now, chatgpt can remix the 250,000 pre-existing otter essays on the web into something that’s just different enough to not violate copyright law. Same goes for graphic artists. Need a graphic of a soup can? Some generative AI can now do it. But this isnt taking work away from the best. Not even close.

Could AI come for my job? Sure, maybe. And I’ll be angry if it does. But I’ll find something else to do.
This.

We’re not talking about Shakespeare here, we’re talking about people who write a product blurb for companies that are just before the point of being so cheap they’ll machine translate something from Chinese.

Companies who can charge more than bottom dollar for their work aren’t at threat, it’s those who cant even get a job there who are threatened.

I bought a bag of ice from Marks and Spencers a while back, there was a tag line on there (because there has to be a tag line on a bag of freaking i

Ah, the rise of AI in the freelancing world – where a neural network’s musings become the bane of the creative class. It’s like HAL 9000 decided to take a detour from space odysseys to dabble in copywriting and graphic design. Remember the good old days when the most complex thing an AI did was lose at chess? Those were quaint times. Now, our silicon friends are not only out-writing the Shakespeares of the ad world but also cutting their paychecks. It’s like watching a real-life episode of Black Mirror, but the twist is that the AI is just really good at Photoshop and persuasive writing.
It’s only good in your mind and imagination. AI can’t create anything new it can only generate similar content to existing content. It is great at mimicking existing content, even blending different types of content, but cannot create something new. It doesn’t know what it is doing. If it writes a story about a mermaid, it doesn’t know what a mermaid is. It only knows things others have written about mermaids.

This study is interesting, but I am skeptical things will remain this way. We’ve been thr
1) LLMs are not AI. They’re a piece of an AI.
2) Even LLMs can be “creative”. The developers put lots of work into keeping it from being TOO creative. They just don’t have “good taste” in what they do. (“Good taste” means they don’t have an underlying model of what they do that goes beyond statistical inference from particle frequency.)
An LLM would be quite unlikely to emit “The gostack distimms the doshes!”, because they haven’t seen those parts. That sentence was created by an author in full intent of
Your definition of “creative” is completely different from what it means to a human. Have you been replaced by a chatbot already?
>It’s only good in your mind and imagination. AI can’t create anything new it can only generate similar content to existing content. It is great at mimicking existing content, even blending different types of content, but cannot create something new. It doesn’t know what it is doing. If it writes a story about a mermaid, it doesn’t know what a mermaid is. It only knows things others have written about mermaids.
Unironically this is overwhelming majority of journalist work. Journalists have no clue on most

AI can’t create anything new it can only generate similar content to existing content.
It doesn’t know what it is doing. If it writes a story about a mermaid
Prompt: Write a story about a mermaid
In the vast depths of the ocean, there lived a mermaid named Luna. She was known throughout her underwater kingdom for her mesmerizing beauty and captivating songs. Her tail, adorned with shimmering scales of emerald green and sapphire blue, glistened in the dim light of the ocean floor.

One day, while swimming through a coral reef teeming with colorful fish and sea creatures, she encountered a human diver named Jack. He had ventured too deep into her territory
Majority of freelancers pivoted half a year ago or so. ChatGPT allowed them to write 5-10 articles where previous they could only write one.
There were some interesting interviews on the subject back then, before the “oh my god he’s using AI, quick call all of your contacts and tell them he’s far right hamas loving, trump loving far left anti abortion pro life hater” movement started. Nowadays they no longer talk about it because they don’t want the hassle of the mob.
Problem is that demand for articles rema
The problem being that in fields with limited demand, this means that more people are freed to do something else.
But if they’re incapable of learning, and were there as a nepotism job, they may be in trouble. And freelance prose writers are very much a nepotism job field. It’s all about knowing the correct people and being in their good graces to get more writing gigs. So a lot of these nepo babies are going to likely be in actual trouble.
Another interesting story I’ve read recently was on topic of high ran
but the twist is that the AI is just really good at Photoshop and persuasive writing.
After you’ve been using it a while, you can see everything has a very samey, shallow texture to it. It doesn’t really make good arguments, it sort of smooshes things together from fragments in a convincing way, but after a few paragraphs it’s clear that it’s just sort of a mush of words.
But AI is very cheap at producing grammatical but otherwise shitty writing.
If you can make your point by providing a few lines to an AI, then save everyone’s time and give those few lines to your readers.
But people won’t. They’ll give those lines to AI to spin out into that weird word salad it’s so good at making, and thereby starts the enshittification of another aspect of life.
A recession no government wants to talk about, but every business is reacting to in the form of layoffs and not-hires. Tech sector employment being decimated in recent times. Banks crashing left and right.
Either freelancers are very easily replaced by still-shitty AI, or this is a bit of bullshit and hype generated around an industry that is simply feeling the effects of everything else.
Bad enough AI is a threat to human employment, but let’s not make it worse with speculative hype. Doesn’t help anyone
Admittedly, the timing is tricky.
However, freelance copywriters and graphic designers are among the most likely to get trounced by current generative AI. They are paid by marketing departments to present the most milquetoast material in the most hollow uninformative way, but sound flowery and nice. This is right up the alley of generative AI. Even before generative AI, it was maddening to try to actually learn about a prospective purchase from a company’s own marketing material. Generative AI had nothin

Admittedly, the timing is tricky.
However, freelance copywriters and graphic designers are among the most likely to get trounced by current generative AI.
They’re also the most likely to get trounced by economic downturns, being a) contractors, not employees, and b) disposable.
No government is talking about it, because a “recession” has a definition, and nothing in economic reporting meets that definition. We did not have two quarters of negative GDP growth, which is the definition of a recession.
Yes, tech employment took a hit, which was entirely predictable after the excess hiring and ridiculous hiring practices of tech, and the predictable “return-to-office” meltdown after the pandemic. More of a predictable mass-stupidity event than anything else.
Yes, there were a few bank failures, mostly due to those banks doing stupid (SVB) or illegal (FTX) things. There was not “crashing left and right” by any means.
And then you go on to chide others for making things worse through exaggeration and “speculative hype” which is exactly what the preamble of your post is.

No government is talking about it, because a “recession” has a definition, and nothing in economic reporting meets that definition. We did not have two quarters of negative GDP growth, which is the definition of a recession.
In a world now loaded with not-a-monopoly mega corporations and Greed in capitalism walking around with a Too Big To Fail card to socialize the worst of their fuck-ups, I love how some still believe that a definition really acts as some kind of viable metric anymore. We barely agree on what a ‘woman’ is these days.
Just how big does a damn online bookstore named Amazon need to become before anyone realizes the definition of ‘monopoly’ is now whatever Greed says it is? Will we believe the definition has bee

I love how some still believe that a definition really acts as some kind of viable metric anymore
Anything can be anything if you make up your own definitions.

you lose all credibility when you start redefining things in your own terms and then argue that “b-b-but definitions arent even real!!1!”
That’s OK. Government lost all credibility when it chose to socialize losses via Too Big To Fail while also granting Greed in capitalism the green light to abuse it again. Not that they have the slightest clue or care as to what is a monopoly anymore. Or privacy. Or security.
Denial is a helluva drug, son.
So your argument is that we should just throw the dictionary out the window and just turn everything into demagogic hyperbole and bullshit, because woke-ism did it first?
And then a bunch of whataboutism completely irrelevant to the topic at hand?
Do yourself a favor – don’t use incorrect terms that have precise definitions to an audience that is predisposed to using precise definitions and accuracy of language when expressing ideas, such as the technicians, engineers, architects, designers, lawyers, and othe
Do yourself a favor and tell me exactly where Too Big To Fail has ever fit into the standard definition of capitalism and the concept of bankruptcy. You really think companies filing for IPOs with valuations in the billions while racking up millions in losses and bragging how they’ve never turned a profit really makes sense?
Perhaps we should raise interest rates another few basis points for ignorance to finally GET it; debt isn’t a good thing. Just ask the college-educated gig worker begging for a college

No government is talking about it, because a “recession” has a definition, and nothing in economic reporting meets that definition.
No government, or anyone else, is talking about how obvious it has become that definition is broken, either.
And nobody will.
Sure, that’s a legit discussion that economics professors, career economists, and banking executives can hash out until they all are blue in the face and come up with something which is generally agreed-to by the global financial community.
But until that happens, we’ll just use the existing agreed-to definition we have which everyone who is slightly knowledgeable about global macroeconomics expects, mmkay?
“You are both eating shit for free, but at least the GDP has increased by two million.”
When the proxy becomes the metric, it becomes worthless.

Banks crashing left and right.
Left and right? This ain’t nothing compared to 2008. Or 1929. Or the 80s S&L implosion.

Bad enough AI is a threat to human employment, but let’s not make it worse with speculative hype. Doesn’t help anyone other than those selling fucking clickbait.
Which is to say, the entire “news” industry, literally every last one. Because news is a business, and it’s not the business of selling news, it’s the business of selling advertising. And the only way to sell advertising in the current market is to scream “THE WORLD IS COMING TO END END!!! LET US SELL YOUR EYEBALLS OR YOU’LL DIE!!! DOGS AND CATS WILL BE LIVING TOGETHER, HELLFIRE WILL RAIN DOWN FROM THE HEAVENS AND SOMEONE WILL KICK YOUR CHILDREN!!! IF YOU DON’T READ OUR NEWS YOU’LL DIE!!!”
Because that’s all
Being in IT design myself, I am making the bots that replace certain types of work.
I enjoy every day of it, because just like any other tool I have built over the last more than 30 years, it doesn’t take away from me, it gives me (and the people I give technology to) the freedom to be MORE creative, to be MORE productive, to be MORE profitable.
As such AI is a great tool to help my developers work faster, since they now have a helpful bot to let them know when they are making mistakes that will one day be ex
I’m not sure where you’re getting your statistics, but the reality is that the economy, and the tech sector, are growing, not shrinking.
Tech sector jobs: https://www.zippia.com/advice/… [zippia.com]
Unemployment rate overall: https://www.bls.gov/charts/emp… [bls.gov]
Sure, big companies like Amazon, Facebook, and Google have laid off some people, but the job market has been so hot that basically anyone who wants a job can get one. “Big Tech” does not equal the “tech sector.”
Generative AI has already cemented a place in illustration work, though I would caution anyone thinking of using it for commercial purposes. It’s valuable though both as a way to communicate your needs to a human illustrator, and it’s awesome for prototyping.
But I haven’t seen any evidence that AI is anywhere close to replacing a graphic artist…yet. All it can do is create web graphics. Even if all you’re doing is web content, it’s still not great at it. A real GD has to go through and correct the results
To me so much of this reads as a perception and marketing issue.
The AI tools loudest proponents that I see out there talking about these things are constantly framing this as a machine to replace artists (thus the term that someone is a ..ugh “prompt artist”) or to let non-trained artists do the artists work when in reality these are really tools for artists to use and save time, like you said for prototyping, fast storyboarding, content-aware-fills, etc. New icons on a Photoshop tool pallet.
If we saw more of the latter rather than the former the sort of uncanny valley disgust with the tech would not be so prevalent. The AI booster people are their own worst enemy when it comes to the PR for this stuff.
How much of the work was in getting a good prototype? It could be not that great at anything else and still severely impact employment.
It cost of labor was the problem there’d be an AI CEO.
If compensation reflected the difficulty, stress, and stakes of the job, this asshole [cnbc.com] wouldn’t make 500 times what this asshole [wikipedia.org] makes.
Neither could much of anything else. Everything that people have depends on an IMMENSE backlog of cultural innovations, most of which almost nobody remembers. Once upon a time all programming was done in assembler.
(And some of which nobody remembers. E.g. “How do you hunt an elephant armed only with a stone knife and a spear with a sharpened wooden point?” As far as I have been able to determine, nobody knows how to do that anymore. Some people did as recently as 2 centuries ago, perhaps more recently.
That’s a valid objection, but the proper answer is to come up with a way to prevent monopolization.
You do realize that movies are stories, right?
After you’ve used AI for a while, you will see clearly that it’s not taking over *anything.* It’s really nothing more than a fancy auto-suggestion that tries to complete your sentences for you while you type your emails.
This article is going to get quoted widely because it confirms people’s preconceptions.
Yet the amount of reduction claimed in the paper was only 2%.
And the data doesn’t show any evidence for what the fluctuations might have been more than two months before ChatGPT. They make no attempt to connect with any other correlative factors besides ChatGPT. For example, is there usually an uptick of freelance work in the months before Christmas, followed by an inevitable downturn? What were the monthly variations i
There may be more comments in this discussion. Without JavaScript enabled, you might want to turn on Classic Discussion System in your preferences instead.
US Debt Interest Bill Rockets Past a Cool $1 Trillion a Year
Microsoft Briefly Restricted Employee Access To OpenAI’s ChatGPT, Citing Security Concerns
“Wish not to seem, but to be, the best.” — Aeschylus

source