A new era of internet spam: How AI slop is taking over our social media feeds

When ChatGPT entered the internet scene in 2022, Tashnuva Sumaiya Islam was working at a digital marketing agency as an executive. At that same time, the agency was struggling to find a skilled copywriter. However, when deliverables were on the line during this transition period, many of Tashnuva's colleagues turned to AI to fill that void.

"The writing was lifeless, to say the least. The copies sounded very obviously AI-written. ChatGPT churned out terms like 'effervescent', 'translucent', 'transcend' and so on, and they copied and pasted those words," shared Tashnuva. 

At the beginning, the agency used AI-written copies word for word. Later down the line, it merely tweaked the copies generated by ChatGPT. 

However, while some may be quick to criticise, it is important to recognise that this approach has become a widespread practice among companies across the world. 

The result? In return for the affordable convenience AI provides us, our feeds and search results have filled up to the brim with passages garnished with the same set of vocabulary and fever-dream-like images.

The internet has reached a new age, and it is filled with spam, or with what the internet calls it 'AI slop'.

If you search the term, you will come across the image of "Shrimp Jesus" and it looks as silly as it sounds. The image features Jesus Christ but with the body of a shrimp, floating underwater. 

Similar nonsensical AI-generated images have been posted online for quite some time as a way of engagement hacking. Not only can you easily make clickbait content, you can probably make your next viral post in a matter of seconds.

The distorted image of Jesus came to represent the surge in misleading and meaningless content that we have become all too familiar with today. 

AI slop is not limited to just visual content either; it has taken over writing as well.

Many English writers have consciously or unconsciously begun changing the way they write. They have particularly begun to avoid words such as "blend," "layered," "dynamic," and so on. Many writers have even begun to avoid em dashes as it is a telltale sign of AI-written content. This has created a separation in human and AI language. 

"In my professional life, I've received many writing samples that were AI-written. This issue has taken over many content-related jobs, and AI has changed the game entirely," commented Tanzim Rahman, senior manager of Digital Communications at Brac International and author of Bangla books such as 'Brishtir Dine Bhara Beshi'.

Interestingly, writing with flaws has become more appreciated, as it is proof that there is a human behind the keyboard. 

But what is it about AI that makes its language seem so lifeless and inhuman? 

AI tools such as ChatGPT create sentences with low perplexity. In fact, it assesses whether a writing sample is AI-written based on how random or unpredictable the language is. 

However, when real people write a piece, it is usually deemed to have high perplexity — it is chaotic and completely random. 

When people write or create art, whether we like it or not, their subconscious thoughts find a way to the canvas. Human writing too is intentional and a reflection of the writer, that is why poetry is analysed word by word. 

Now, imagine all of that depth and complexity just vanishing — that is how AI slop fills our feeds. 

"A problem I see in the coming years is that AI will write in a more human way. It will borrow language styles from classic literature, and it will refine its own writing style," commented Tanzim.

He continues, "Many writers are contemplating how to refine their writing style so that it does not look like AI. It is fun to imagine a scenario where writers retreat to more minimalistic styles like McCarthy or Ernest Hemingway."

In terms of writing, the creativity has actually shifted from the page to the prompt. 

When the AI action figure trend surfaced on social media, many users included the prompt in their posts so that others could use it too. The trend consisted of AI-generated images of the user as an action figure in its packaging, with other details that add to the action figure's character. 

Users included the prompts as if they were the password to a hidden feature on your device, and they were doing their due diligence to make it accessible. 

This also reflects how people still have an appreciation for art, especially if it feels personal to them. However, they prefer the convenience and affordability of artificial intelligence. 

Moreover, AI-generated art and writing have infiltrated our consciousness more than we realise. 

Nazneen Meem, a freelance artist and the founder of Wooden Dreams, shared how AI has impacted her business. 

"At times, clients provide reference pieces that are AI-generated. When people approach me for artwork, it is usually because they feel like my style agrees with their brand image," shared Meem. "However, when you bring AI into the mix, it takes all the creativity out of it, and my style ultimately won't be translated in the end product."

The AI generated samples are conjured from the "mind" of the tool — or at least the collection of information that it has. 

Furthermore, according to Meem, people usually commission logos with AI samples as they want the rights to the image, in which case, an artist's creativity is entirely replaced by generic AI-generated content. 

Though AI is in its infancy, it is already replacing skills from creative professionals, leading to a more concerning question of what will happen in the long run. 

According to a research conducted by Michael Gerlich, published on Societies, AI could diminish our critical thinking skills. 

"Younger participants exhibited higher dependence on AI tools and lower critical thinking scores compared to older participants. Furthermore, higher educational attainment was associated with better critical thinking skills, regardless of AI usage," he wrote. "These results highlight the potential cognitive costs of AI tool reliance, emphasising the need for educational strategies that promote critical engagement with AI technologies."

Tanzim also believes that AI could be a major culprit of the spread of misinformation, largely due to its inability to say, 'I don't know'.

"When you enter your query, it is common to see faulty search results appear in AI tools, especially in ChatGPT and Google's Gemini. ChatGPT eventually had to provide a disclaimer for their answers. The fact that AI tools have to provide an answer even if it doesn't have it is a huge problem," he added.

But the truth is, AI is getting better day by day,

A few months ago, AI could not draw hands perfectly, and distorted fingers became the telltale sign of AI art. However, now it has fixed that problem and as it evolves, it will become more and more difficult to tell AI apart from human-made art.

Cartoonist Mahatab Rashid believes in creating a separation between human creativity and AI.

"Artists are artists, but AI still has a place in art. It can help with menial, time-consuming labour that doesn't include decision-making or cerebral tasks," said Mahatab. 

This is not the first time technology has created a wave of fear among people either.

Tanzim recalled how Photoshop was first received by the public, and how it was misused to change how people view the truth, and that it was ultimately a "lazy shortcut".

"I see a lot of those arguments being repeated for AI. On a fundamental level, I don't think AI is bad. To me, art is an expression. If that expression comes from a human being — be it with Photoshop, a laptop, a rock, or AI — it is still human art," shared Tanzim.