AI Slop, Advertising and Visual Literacy

B

benc98

Messages
1,218
Name
Ben
Edit My Images
Yes
I was a bit surprised to be served an Ad on Threads yesterday, the image was textbook AI Slop and advertising a series of crime novels available to purchase on Amazon.
I saw another chart somewhere that showed the increase in books on Amazon since AI has taken off - I'm not saying the book I saw advertised was written by AI, but I made the connection and wouldn't buy it.

copilot.microsoft.com

Your poster is ready now — gritty, moody, very British, and absolutely in the spirit of a DCI Parker series launch....

Shared via Copilot
copilot.microsoft.com copilot.microsoft.com

And then on Facebook I see a pub near me advertising their Sunday lunch menu, the image showing four dinners with different roast meats - again what I recognise as a very typical AI image. A commenter on the post asked if the photos were fake, and poster responded no and said they had been busy in the kitchen. The photos looked too perfect, shared the same composition and that saturated AI look.


copilot.microsoft.com

Your image is ready now — a warm, cosy, traditional Sunday Roast poster for **The New Inn**, featuring all four roasts, Yorkshire puds, gravy, veg, puddings,...

Shared via Copilot
copilot.microsoft.com copilot.microsoft.com

I spent no more than 3 minutes making my versions and look no worse than what I was served

I've also noticed lots of AI generated ads for Events on social media - https://copilot.microsoft.com/shares/fPXu6U3v1RyBsHMP22RM5

While it's great that AI has democratised graphic design, as a consumer I feel these adverts are a massive turn off for me (connotes inauthenticity, fake, etc..) and I feel less likely to visit the pub, buy the book or visit the event. I wonder if other people are subsequently put off by the services and products being advertised, consciously and subconsciously by people are are less visual literate? Wonder if people will begin to push back against it or just accept it.. toughts?
 
Look out for humans deliberately introducing 'flaws' into their work to differentiate it.
 
benc98 said:
While it's great that AI has democratised graphic design, as a consumer I feel these adverts are a massive turn off for me (connotes inauthenticity, fake, etc..) and I feel less likely to visit the pub, buy the book or visit the event. I wonder if other people are subsequently put off by the services and products being advertised, consciously and subconsciously by people are are less visual literate? Wonder if people will begin to push back against it or just accept it.. thoughts?
Click to expand...

I don't know if I've been deterred exactly, but I just tend to ignore them. I suppose it's the nature of AI that it's evolved to almost a common look, and it makes it very easy to spot. It is uniquely bland, in my experience.
 
There are local photography groups that use AI to advertise for new members on Insta. The amount of comments panning them each time they're posted always makes me smile.

Thankfully there are still enough of us around not afraid to call it out where required.
 
It instantly puts me off, it's unauthentic and makes me wonder what other corners they cut as a business.

That generic AI font really irks me as well.
 
benc98 said:
I was a bit surprised to be served an Ad on Threads yesterday, the image was textbook AI Slop and advertising a series of crime novels available to purchase on Amazon.
I saw another chart somewhere that showed the increase in books on Amazon since AI has taken off - I'm not saying the book I saw advertised was written by AI, but I made the connection and wouldn't buy it.

copilot.microsoft.com

Your poster is ready now — gritty, moody, very British, and absolutely in the spirit of a DCI Parker series launch....

Shared via Copilot
copilot.microsoft.com copilot.microsoft.com

And then on Facebook I see a pub near me advertising their Sunday lunch menu, the image showing four dinners with different roast meats - again what I recognise as a very typical AI image. A commenter on the post asked if the photos were fake, and poster responded no and said they had been busy in the kitchen. The photos looked too perfect, shared the same composition and that saturated AI look.


copilot.microsoft.com

Your image is ready now — a warm, cosy, traditional Sunday Roast poster for **The New Inn**, featuring all four roasts, Yorkshire puds, gravy, veg, puddings,...

Shared via Copilot
copilot.microsoft.com copilot.microsoft.com

I spent no more than 3 minutes making my versions and look no worse than what I was served

I've also noticed lots of AI generated ads for Events on social media - https://copilot.microsoft.com/shares/fPXu6U3v1RyBsHMP22RM5

While it's great that AI has democratised graphic design, as a consumer I feel these adverts are a massive turn off for me (connotes inauthenticity, fake, etc..) and I feel less likely to visit the pub, buy the book or visit the event. I wonder if other people are subsequently put off by the services and products being advertised, consciously and subconsciously by people who are less visual literate? Wonder if people will begin to push back against it or just accept it.. toughts?
Click to expand...
People "are" pushing back against it and It's been discussed a few times before.

There is increasing evidence that people are very good at recognising AI slop, and are put off buying the product because they see a seller using AI images as "untrustworthy".

High-end commercial photographers are busier than they have ever been. Scott Choucino (https://scottchoucino.com/) has commented that he and others in the industry have never been busier as clients want to distance themselves from AI-generated photographs. He also mentions that while there have always been some commercial photographers who never switched to digital, the level of interest in having advertising campaigns shot on film is showing a resurgence, just to make it more obvious that the pictures haven't been generated by AI.

For me, I think AI is probably the most dangerous thing human beings have ever invented, and as the already rich and powerful are going to get even richer and more powerful on the back of AI, I fear it's unstoppable.
 
myotis said:
For me, I think AI is probably the most dangerous thing human beings have ever invented, and as the already rich and powerful are going to get even richer and more powerful on the back of AI, I fear it's unstoppable.
Click to expand...
Second most dangerous in my opinion.

Lawyers are still in the lead by a short head! :exit:
 
To add to my earlier post, Scott Chocuino has just been discussing the amount of money that large advertisers are spending on proving their ads aren't AI.

For example, routinely making Behind the Scenes (BTS) videos of how the advert was made for them to distribute with the advertisement.

It's a long rambling video but the relevant part is at 22:51 minutes

View: https://www.youtube.com/watch?v=UAfU8tj2OC4


So folks, as customers, we aren't only paying for the adverts, but also the advert for the adverts to prove they aren't AI.

No doubt we will soon be paying even more to have it proved, that the "proof" of something not being AI, wasn't also produced by AI !
 
myotis said:
No doubt we will soon be paying even more to have it proved, that the "proof" of something not being AI, wasn't also produced by AI !
Click to expand...
That's all right: we just set up our AI to check their AI.

Of course, this could end up with an endless chain of AIs checking one another, which will draw so much power that there'll be none left for us to turn on our computers to look at anything...

:tumbleweed:
 
AndrewFlannigan said:
That's all right: we just set up our AI to check their AI.

Of course, this could end up with an endless chain of AIs checking one another, which will draw so much power that there'll be none left for us to turn on our computers to look at anything...

:tumbleweed:
Click to expand...
But I'm not sure the AI is doing anything more than collating data at a rate faster than we could do as humans.

So the misinformation generated by the first lot of AI, is simply being recycled by the second AI. It needs human beings to keep feeding it with new, original, verified information.

It's a fantastic tool in the right hands and used for the right application, but it looks as if it is going to soak up all the world's power, and all the world's RAM.
 
myotis said:
It needs human beings to keep feeding it with new, original, verified information.
Click to expand...
Aha! the old GIGO (garbage in, garbage out)
juggler said:
Convolutional Neural Networks are really good at that stuff.
Click to expand...
Well, the current crop of humans seem to disprove that theory. Any calculation of any form is only as good as the base assumptions from which it starts ... and on the evidence so far, the base assumptions are largely rubbish.
 
juggler said:
Amusingly.. it's industries like law which are likely to be most affected by AI. Specialist AI models, not genAI stuff. A lot of law is about recalling & synthesising huge quantities of facts & arguments. Convolutional Neural Networks are really good at that stuff.
Click to expand...
I think this applies to other professions as well, but I feel the issue is that a lot of a persons initial learning, building up an "intuition" on how things work comes from working through these sorts of tasks.

My fear is that we end up with a new generation of "experts" which have no foundation to their expertise. I'm also worried in the loss of jobs, if "trainees" are replaced with AI, so we end up with no one getting trained, and forced to rely more and more on AI as some sort of oracle, which seems to be tremendously flawed without proper human oversight.
 
Last edited:
I'm seeing slop more and more in advertisements. There are entire youtube channels that host nothing but AI created and AI narrated videos, and they have thousands of subs. They have the same style as written AI, repeating the same points several times just worded differently, then bolting on a little phrase to button the whole thing up - "When you think about it, there's something deeply profound at the core of that thought" or some other tosh.

I have grown to hate it with a passion.

Don't get me wrong, AI in general is great. I use it almost daily both in and out of work. It's great for coding (lol), logic checks, tough calculations, data aggregation, looking for trends etc.

The quote "I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do laundry and dishes." often gets picked apart by AI bros with arguments like "hurr durr AI is limited to digital content not physical tasks" or "robots that do physical stuff are too much of a liabilty" etc.
The sentiment isn't to be taken literally, it just means use the technology in an appropriate way. Let it deal with facts, numbers, calculations and things that save you time. For anything that requires creativity, original thought, or an opinion - It is incapable of any of those, and rather than admitting it, it will just give you slop.
 
jimmyjamjojo said:
...It's great for coding (lol), logic checks, tough calculations, data aggregation, looking for trends etc.
Click to expand...
I'm glad I left the IT industry ten years ago.

The idea that I would be expected to entrust logic checking to a program that I couldn't check myself, makes me wonder if the the idiot managers have finally succeeded in turning the office into a madhouse.

:thinking:
 
Had to renew my insurance over the phone this morning ... obvious AI assistant answered the phone...

"To renew without changes press 1, to renew with changes press 2".

Presses 2 ... "You have made an incorrect choice!"
 
Claire, the AI assistant at EonNext, pays my electricity bill for me. Make of that what you will.
 
myotis said:
I think this applies to other professions as well, but I feel the issue is that a lot of a persons initial learning, building up an "intuition" on how things work comes from working through these sorts of tasks.

My fear is that we end up with a new generation of "experts" which have no foundation to their expertise. I'm also worried in the loss of jobs, if "trainees" are replaced with AI, so we end up with no one getting trained, and forced to rely more and more on AI as some sort of oracle, which seems to be tremendously flawed without proper human oversight.
Click to expand...
I have mixed feelings about that. In my own (former) profession - software engineering - there's a lot of talk of hordes of developers being replaced by AI coding agents.

IME most software engineers are doing nothing of value or with any longevity. AI is doing the same meaningless churn that the majority of developers do. Most of the work in software is done by a small fraction of talented people; AI is just making that more obvious.

Maybe other industries are the same? I have no idea.
 
juggler said:
Most of the work in software is done by a small fraction of talented people; AI is just making that more obvious.
Click to expand...
That wasn't my experience in the thirty plus years I was at the sharp end.

I always found that producing robust software was a collaborative process, from identifying a need all the way to deployment and daily use. Some systems were small but even then you needed a sponsor on the business side to identify the need, all the way to testers and trainers, to get the users ready for deployment. Of course, this was the business user case and other models of development exist.
 
BBC Panorama 7:30pm
Cute, funny, and sometimes surreal, AI-generated videos are flooding our social media feeds, but behind the dancing dogs and talking fruit lies a far darker world. In this Panorama investigation, reporter Marianna Spring explores how AI slop is shaping what we see, what we believe and who we can trust. From scam adverts and cloned voices to deepfake fraud, political manipulation and racially charged disinformation, the programme reveals how cheap and easily available AI tools are being weaponised. We ask, in the age of artificial intelligence, will seeing ever be believing again?
Click to expand...
 
juggler said:
I have mixed feelings about that. In my own (former) profession - software engineering - there's a lot of talk of hordes of developers being replaced by AI coding agents.

IME most software engineers are doing nothing of value or with any longevity. AI is doing the same meaningless churn that the majority of developers do. Most of the work in software is done by a small fraction of talented people; AI is just making that more obvious.

Maybe other industries are the same? I have no idea.
Click to expand...
Wearing my statistician's hat, AI seems like a great tool for dealing with the drudgery of collating and cleaning data, but in reality, that drudgery was actually a period of discovery and insight. So before you began any analysis, you had a deep understanding of the data you were dealing with. It would highlight anomalies or oddities that needed explaining before you could proceed.

Sometimes, these would be so great that they highlighted major flaws in the data, making it worthless.

It's a stage that no one enjoys doing, and people (students and professional scientists) would often skip doing it, and it was only when they asked for my help at a later stage, and I would have a look at the raw data and find flaws in the data, that meant all the work they had done before asking for help had been wasted.

Although most people think of statistical analysis as a mechanistic crunching of numbers, it isn't because over time you develop an intuition for when things just don't feel right. It also requires insight not only from expert statisticians but also from experts in the field the data was collected .

It's important that statistical analysis doesn't just become a "black box" where you throw numbers in one end and an answer comes out the other, because using statistics to make good decisions (which is what it's all about) requires you to understand what happens inside the black box and to what extent that should affect your decision.

Even before AI, there were recurring concerns from professional statistician, of problems with "easy to use" statistics software which gave invalid results. But they were regularly used in research because of their ease of use.

I'm not saying AI hasn't got value, because it has, but I feel we need to think carefully about what we risk losing if we don't control its use now.

On the programming front, I am coming across an increasing number of people who, with the aid of AI, and no knowledge of programming, are making small specialist programs for dedicated purposes. Which is great., it's a bit like smartphones with cameras opening up photography to many people who might otherwise not enjoyed photography.

But the concern raised in my earlier post still exists: how do these small number of talented programmers (that are still important) learn their trade, if they don't go through the churning of code phase, which allows them to develop an "intuition" for code and coding.

That is certainly what is missing from scientists who only use "easy to use" stats software, and that will get worse as AI use increases. Who is going to realise AI is churning out rubbish if they don't know how to recognise it.
 
Last edited:
AndrewFlannigan said:
That wasn't my experience in the thirty plus years I was at the sharp end.

I always found that producing robust software was a collaborative process, from identifying a need all the way to deployment and daily use. Some systems were small but even then you needed a sponsor on the business side to identify the need, all the way to testers and trainers, to get the users ready for deployment. Of course, this was the business user case and other models of development exist.
Click to expand...
I think this applies to lots of professions, that people see AI replacing or at least dramatically reducing staff numbers.
 
myotis said:
I think this applies to lots of professions, that people see AI replacing or at least dramatically reducing staff numbers.
Click to expand...
When this happens, where is the money going to come from? If loads of people are out of work, it will be the senior management (for a short period of time) and the shareholders that benefit from AI reducing staff numbers.

Community, Society and Humanity won't benefit. Unless there is a different way of giving people a standard of living.
 
Pound Coin said:
When this happens, where is the money going to come from? If loads of people are out of work, it will be the senior management (for a short period of time) and the shareholders that benefit from AI reducing staff numbers.

Community, Society and Humanity won't benefit. Unless there is a different way of giving people a standard of living.
Click to expand...
I think the theory goes that AI will be taxed, and that will pay for us all to have universal basic income. A mad theory if you ask me
 
Pound Coin said:
When this happens, where is the money going to come from? If loads of people are out of work, it will be the senior management (for a short period of time) and the shareholders that benefit from AI reducing staff numbers.

Community, Society and Humanity won't benefit. Unless there is a different way of giving people a standard of living.
Click to expand...
I often think about this.

We seem to be replacing many manual jobs with robots and machinery (being going on for a long time) and we are now replacing white collar jobs with AI.

With the promise of lots of new jobs, just of a different kind.

The pessimist in me, sees a world with a small number of wealthy people, being served largely by AI and robots with a small population of "real people" being maintained to service the AI and the robots, plus the few real people jobs that still exist.
 
benc98 said:
I think the theory goes that AI will be taxed, and that will pay for us all to have universal basic income. A mad theory if you ask me
Click to expand...
But that is inefficient. The competition is on a World level and companies which are not heavily taxed will get the business. Productivity through the now out of work people has to increase so they are no longer a burden. Or, to put it another way, any dreams of machines working while they pay for everyone else are short sighted.
 
myotis said:
Won't the AI "just leave the country" if we try to tax it.
Click to expand...

If it provides a service that people pay for them it should be possible to tax the service sold here because they're doing business in the UK.
 
AI is useful, certainly, but requires much more care than many think. A company I consult for uses Claude quite a lot, my main employer uses copilot. I've been working with both this morning, seeing contradictory findings from both to try to resolve the way forward on a project. A difficulty is that they can sound utterly convincing and authoritative on a subject when they have no real knowledge. Also they seem to push conversations in unintended directions or bring factors that you don't want to the discussion. Knowing how to question is a big part of managing their output too.
 
You must log in or register to reply here.
Back
Top