Surviving AI Slop

Corporate AI Slop Goes Exponential

Obviously, AI didn’t invent cor­por­ate cringe. But it will indus­tri­al­ise it.

You know the posts.

“I’m humbled to announce…”
“Three les­sons my tod­dler taught me about lead­er­ship…”
“This changed my per­spect­ive…”
“I wasn’t going to share this, but…”
“Here’s what nobody talks about…”

They used to be writ­ten by humans. Carefully. Laboriously. With a faint hope that the algorithm gods would smile upon them. ³Silfwer, J. (2020, May 23). Corporate Cringe. Doctor Spin | The PR Blog. https://​doc​tor​spin​.net/​c​o​r​p​o​r​a​t​e​-​c​r​i​n​ge/

Now, that abnox­ious brand of LinkedIn slop is gen­er­ated at scale.

We recog­nise the struc­ture. Short para­graphs. Dramatic spa­cing. Vulnerability. Redemption. A numbered list. A mor­al. A hand­shake emoji. A rock­et. A call to engage.

Call me a PR snob, but:
Yuck.

The fas­cin­at­ing part is that AI slop doesn’t neces­sar­ily make these posts worse. It makes them ubi­quit­ous. And ubi­quity is what turns mildly annoy­ing into aggress­ively cringy.

The tone. The cadence. The fake humil­ity. The tem­plated vul­ner­ab­il­ity. The safe les­sons. The algorithm-friendly format­ting. It was all struc­tured for engage­ment long before AI entered the chat.

Once you’ve seen one “I got rejec­ted 47 times before suc­cess” post, you’ll find it inspir­ing.
After the hun­dredth, it’s a par­ody.
After the thou­sandth, it’s back­ground noise.

AI did­n’t invent cor­por­ate cringe; it scales it. AI simply removed the last bot­tle­neck — human effort.

Now the feed fills with per­fectly accept­able, pro­fes­sion­ally struc­tured, emo­tion­ally cal­ib­rated posts that all feel… identic­al. Nobody is say­ing any­thing wrong. And nobody truly listens.

Which is how we arrive at a new cat­egory of con­tent:
Corporate AI slop.

Not offens­ive. Not stu­pid. Not even badly writ­ten.
Just aggress­ively average.

The Signal-to-Noise Collapse

AI evan­gel­ists are eager to point out that the slop­pi­ness of today’s AI-gen­er­ated con­tent will nev­er get any worse.

But the prob­lem isn’t con­tent qual­ity; it’s inform­a­tion over­load.
AI slop isn’t the same as bad con­tent; it’s excess­ive content.

AI slop doesn’t lower qual­ity — it bur­ies it.

As AI slop is becom­ing the new stand­ard of men­tally drown­ing ourselves, the only tan­gible solu­tion is to throw AI gen­er­at­ors at the algorithms them­selves, allow­ing them to syn­thes­ise excess­ive AI slop into… more con­densed forms of AI slop, I guess.

Clay Shirky fam­ously stated, “There’s no inform­a­tion over­load, only fil­ter failure.” 

Today, we’re a few dec­ades deep into the social media revolu­tion, and we still haven’t solved the issue of our lim­ited men­tal band­width. If the sig­nal-to-noise ratio was strained before, AI gen­er­at­ors are now enter­ing the chat with fuck­ing bulldozers.

If any­thing, as a res­ult of user-gen­er­ated con­tent and social media feeds, we have so far placed the keys to our pre­cious atten­tion spans in the hands of a few cent­ral­ised algorithms. 

And we aren’t allowed to peek inside these algorithms, either.

We. Hate. The. Algorithms.

I can­’t help draw­ing par­al­lels with social media. All that demo­crat­isa­tion of con­tent pub­lish­ing (“Here comes every­body”), yet we still hate those algorithms.

We. Hate. The. Algorithms.

Yes, “hate” is a strong word. But after being a digit­al PR spe­cial­ist since 2005, I must call a spade a spade at this point. The algorithms draw us in. We’re at their mercy. We know it, and we hate it. But we can­not stay away.

Earning trust over time gets the algorithmic back­seat, while sen­sa­tion­al pieces of con­tent get the roy­al treatment.

Do we hate algorithms because they aren’t good enough?
Or do we hate them for being too good?

Our men­tal band­width is bio­lo­gic­ally lim­ited, which is why algorithms are so power­ful to begin with, and so we can­’t really out­source all of our media con­sump­tion to machines of our own. Sure, I could ask an AI agent to read The New York Times, watch Breaking Bad, or play Red Dead Redemption 2 on my behalf, but what would be the point?

We must con­sider the pos­sib­il­ity that we’ll con­tin­ue to hate algorithms — des­pite them being powered by AI.

Enter: Artificial Spin Doctors

If the Margaux Blanchard incid­ent was an indic­a­tion, we’re stand­ing on the pre­cip­ice of autonom­ous PR bots pitch­ing the media en masse. ⁴Silfwer, J. (2026, April 2). The Margaux Blanchard Incident from a PR Perspective. Doctor Spin | The PR Blog. https://​doc​tor​spin​.net/​m​a​r​g​a​u​x​-​b​l​a​n​c​h​a​r​d​-​i​n​c​i​d​e​nt/

Enter: The age of arti­fi­cial spin doctors.

I can abso­lutely see how this will work.
And — I can also see how it won’t.

(I’m actu­ally in the pro­cess of pour­ing my dec­ades of exper­i­ence into build­ing an exper­i­ment­al and fully autonom­ous PR agent for industry research, media lists, and pub­li­city out­reach as we speak.)

While I’m not too sure about a pos­it­ive out­come, I do know that we will try these arti­fi­cial spin doc­tors out. Even if you and I decide not to go down that route, oth­ers will, and even­tu­ally we must fol­low suit. That’s just how tech­no­lo­gic­al adop­tion works.

Journalism will have no oth­er choice but to respond in kind.

To think that arti­fi­cial journ­al­ists will inter­act with arti­fi­cial spin doc­tors is kind of wild. 

Autonomous AI agents on both sides will be trained on essen­tially the same data­sets. Industrial-scale story gen­er­a­tion and pitch­ing to be pushed through edit­or­i­al sort­ing machines, cater­ing to per­son­al­ised news­feed algorithms through instant­an­eous media cycles.

Wild, indeed.

How Industrialisation Solved for Slop

We’ve seen mass-pro­duc­tion at play before:

The Industrial Revolution didn’t just mass-pro­duce goods. It mass-pro­duced mediocrity. Perfectly func­tion­al chairs. Perfectly accept­able clothes. Perfectly adequate watches. Nothing wrong with any of it — just an ava­lanche of sameness.

When fur­niture factor­ies star­ted churn­ing out identic­al tables, crafts­man­ship didn’t dis­ap­pear. Handmade fur­niture became a sig­nal.

The same thing happened to watches. Quartz move­ments nearly wiped out mech­an­ic­al watch­mak­ing in the 1970s. Cheap, accur­ate, mass-pro­duced timepieces flooded the mar­ket. Mechanical watches sur­vived not because they were bet­ter at telling time — they weren’t! — but because they became expres­sions of taste, engin­eer­ing, and human effort.

Vinyl records weren’t killed by stream­ing. Tailoring wasn’t killed by fast fash­ion. Craft beer didn’t dis­ap­pear when indus­tri­al lager conquered the world. In each case, indus­tri­al­isa­tion cre­ated abund­ance. Abundance des­troyed per­ceived value. Craftsmanship became scarce. 

Scarcity cre­ated… sig­nal.

AI slop is the mass pro­duc­tion phase. Functional blog posts. Functional LinkedIn updates. Functional press releases. All tech­nic­ally fine. But also some­what inter­change­able. Mediocre, even.

The more AI gen­er­ates safe, pol­ished, aver­age con­tent, the more we start noti­cing the noth­ing­ness. When everything sounds pro­fes­sion­al, pro­fes­sion­al­ism stops sig­nalling anything.

This is why AI slop will struggle to erad­ic­ate crafts­man­ship. Handmade con­tent doesn’t have to be bet­ter to be more valu­able. When every­one can mass-pro­duce high-qual­ity con­tent, taste becomes the dif­fer­en­ti­at­or. Judgement becomes a skill. Voice becomes the moat.

Imperfection becomes signal.

How PR Will Survive AI Slop

We will sur­vive the com­ing onslaught of AI slop and AI enshittification.

Not by com­pet­ing on volume. Machines will always pro­duce more.
Not by com­pet­ing on pol­ish. Machines will always be clean­er.
Not by com­pet­ing on speed. Machines will always be faster.

Those battles are already lost.

Instead, the com­pet­i­tion shifts to per­spect­ive. Judgement. Originality. Courage. Personality. Art. The things that don’t emerge from pre­dic­tion engines trained on aver­ages. The things that require taste.

This is where the future dif­fer­en­ti­at­ors start to become obvious.

Real rela­tion­ships.
Real expert­ise.
Real taste.
Real voice.
Real risk-tak­ing.
Real imper­fec­tion.

Sure, these qual­it­ies don’t scale par­tic­u­larly well. They don’t lend them­selves to tem­plates. They don’t emerge from prompt­ing. And that’s pre­cisely why they will mat­ter more.

The more AI slop and plat­form enshit­ti­fic­a­tion we get bom­barded with, the more we value human con­nec­tion. The more pol­ished everything becomes, the more we appre­ci­ate those rough edges. The more aver­age everything sounds, the more we listen for a voice — flaws and all.

That’s the opportunity.

Not to fight AI. Not to reject auto­ma­tion. Not to pre­tend the slop isn’t com­ing. But to lean into the things that AI struggles to instil in oth­er human beings — con­nec­tion, per­son­al­ity, taste, and the cour­age to sound like yourself.

Those are the build­ing blocks of human relationships.

And lo and behold, this is not too far off from what PR is sup­posed to be about any­way. Right?