We need to talk about "slop"

What Zohran Mamdani, AI, and Cluely have in common

“Slop” no longer means low-quality content. Most slop we see today is not crappy SEO pages, listicles on Buzzfeed, offshore content farms, or even bots on X or Reddit.

We need to update both our definition and mental models.

“Slop” was always about a lack of substance, a lack of care.

A lack of care used to correlate with a lack of production value, taste, and quality. But it is now possible to make a high volume of things with high “production value,” but zero care.

Which means that today, the most dangerous form of slop is the kind that doesn’t look like it: things with low care / substance, that are extremely polished or even tasteful.

Why does this matter?

Slop is high-fructose corn syrup: addictive, filling, and nutritionally empty. It’s tasty in the moment, but is it worth obesity and an eviscerated microbiome?

Lack of care and lack of substance degrade society the same way.

Corporations prioritizing shareholder value. American consumer products offshoring production and losing brand value. Mainstream media devolving into clickbait & bias, losing public trust. Politicians prioritizing special interests / lobbyists post-Citizens United. These are all cases where major institutions lost broad societal trust because they prioritized the short-term, low=nutrition, high-calorie hit.

For each individual player, their actions were rational. But that’s prisoner’s dilemma for you: “defecting” — choosing the short-term, local maxima — is the rational choice, but leaves everyone worse off and sacrifices the global maxima.

I’m all for rational, self-interested, high-agency choices. “You can just do things” is very empowering! But… just because you can just do things, doesn’t mean you should. Society is better off when actors make thoughtful holistic decisions that create win-win outcomes: when players are “long-term greedy,” solving for multi-turn games.

It’s Silicon Valley’s turn. We’ve already come dangerously close to losing societal trust due to thoughtless, unintended consequences: tech’s impact on labor markets, social media addiction, Cambridge Analytica. We will start recovering some of it with things like a refocus on American Dynamism, financial freedom, etc. But the real test is going to kick in based on how we handle the arrival of AI, and in big part: what do we do about slop?

Slop is accelerating across domains, thanks to AI

“Thought leaders” can now pose ostensibly sophisticated ideas which are actually ridiculously stupid, and distribute at scale by ruthlessly targeting receptive people across a massive audience.

• This first wave started a decade ago with social media algorithms and SEO, but it really started accelerating a year or two ago thanks to LLMs.

• This has now been brought down to a science by motivated — sometimes malicious — actors.

• At worst, LLMs can also hallucinate apparently interesting things which aren’t real at all. At best, they’re empty calories, with no substance to the content.

Content creation has become sloppified over the last 20 years: step by step, it’s become possible to create ostensibly polished content in just a few minutes

• Algorithmic juicing of viral posts and videos with Facebook, Vine, and Instagram

• The TikTokification of video content, which further shortened attention spans

• The advent of editing tools (especially with AI) the last 2 years — polishing a turd has become an order of magnitude easier

• The latest image and video models (e.g., Veo3) make it easy to blurt out a few words and make something that would have taken a 10-person team a week, even a few years ago

There has been a massive spike in ostensibly high-growth companies with horrendous cracks in the foundation:

1. With the rise in ability to build software really quickly and acquire customers (on both a prosumer subscription and enterprise openness to AI) rapidly, we’ve seen hypergrowth like never before

2. But many of the products have no moat, really bad churn masked by even faster growth (for now), and — worst of all — sketchy metrics.

3. We see some variation of this during every big hype cycle (sharing economy in 2015-18, crypto in 2020-22), but the rate today thanks to AI today is unmatched.

Let’s look at examples of these in politics, AI content, and startup hype.

Political slop = lopsided [style : substance] ratio

You’ve probably seen these videos:

Zohran's policies are incredibly problematic — rent freezes will result in degrading housing stock, city-run grocery stores are a pipe dream (“they won’t have to pay rent?!”) which will probably hurt local small businesses (the main source of groceries in NYC is bodegas, not Walmart).

But his brand has terrific cinematography. Incredible, tasteful production value, and “rizz.” I actually respect that.

Hyper progressive administrations have failed catastrophically in San Francisco and elsewhere. But Zohran’s ideas sure sound nice, and he’s saying them in a softly lit video with a charismatic, empathetic voice. And he just won the democratic primary.

AI slop = power wielded without thoughtfulness or duty of care

I talked to a founder last year who was building a “purely AI TikTok — we won’t be bottlenecked by human creators, so we can create 100x as much content.”

I said, “oh that makes sense, so much of TikTok is algorithmic anyway. I have one request: can you at least make it higher-nutrition rather than the slop on there now?”

He replied, “actually we’re going to focus on the lowest-nutrition stuff because that tends to be easier to produce and is more likely to go viral anyway.” I’ve never encountered a clearer origin story for AI slop.

All technology — from the wheel to the steam engine to the internet to AI — is simply a source of leverage. Its good or bad effects stem mostly from the extent of care exercised by the wielder.

The internet gave people a 10-100x point of leverage. AI is a 10-100x multiplier on top of that. So someone who’s very good at AI, and very good at the internet, is possibly 10,000x as effective as someone who’s good at neither. And when that happens… AI-aided craftsmanship, when paired with taste and care, is a sight to behold.

But just like a 10,000-lb elephant can accidentally squish someone, a human with 100x+ leverage can cause incredible harm. None of this is to say “don’t reach for the holy grail” or gatekeep in any fashion. It’s merely to say the present moment should instill in us — both the builders and the witnesses — an indelible duty of care.

Startup slop = the potent combination of content slop, AI slop, and hype slop

Warp thought they were being clever by giving out sponsor badges on X to affiliates. Turns out one of them was racist (like, really bad).

Series is apparently using their VC capital to rent out a mansion in the Hamptons, hiring “interns” who’ll fly in on first class flights, to participate in an intern reality show, “get lit on weekends”. Their CMO (who’s still in college, of course), promises “i'll bunk bed with my fav applicant.” What could possibly go wrong.

Molly O’Shea recorded a great interview with Roy Lee. His worldview roughly boils down into “you have to do outrageous things to get millions of views,” and “virality is the most important thing in the world” and “if you can do it consistently, you’re untouchable.” (Emphasis mine.)

And then he goes on add this bit:

If I make it, it will be, like, the death of professionalism.

I don’t actually dislike Roy. He seems incredibly sharp and talented, and I want smart founders to win. But things like this, encouraging other founders that professionalism (aka being civilized in the workplace) doesn’t matter, is simply a flippant way of saying “I’m burning down the scaffolding that holds up trust in the ecosystem.” Saying “cheat on everything” sends the wrong message: that founders don’t (and shouldn’t) care about ethics. This worries me. With the immense power of AI just being realized, thoughtfulness and integrity is more important, not less.

Besides, I’m not sure the impulsiveness is in Cluely’s best interest; building a durable company requires forethought and duty of care. If he’s supposed to be like Zuck, then I hope his allies become his Thiel and Sandberg — challenging (not indulging) impulsiveness, and encouraging thoughtful decisions with long-term implications.

What is “good” distribution in the age of slop?

“If a tree falls in the forest and no one hears it, did it even fall?” It may as well not have. Distribution is an essential important force multiplier, and I do think many founders (myself included) underestimate its importance.

The old ways of doing it no longer work:

1. The pace that the world is operating now is no longer conducive to 1:1. We’ve got to go many : 1.

2. There is too much noise, and things like cold email no longer work.

3. There’s now a lot of decent software out there. This is good! The bar is higher! But it makes it harder than ever to differentiate on product.

So what’s left?

I’m by no means an expert. I have all of 7k followers on X. My dog has more followers than me on Instagram. I’ve never downloaded spyware TikTok.

But I do know a good bit about startups and this is my best guess.

Build hard things

Making hard things, in a thoughtful way is harder than ever, in an era with TikTok attention spans keep people from truly finishing and polishing their work.

Link to post →

1. In a startup, this means building hard technical things that can’t be Sherlocked in a couple of weeks by a motivated team

2. In Mr. Beast’s world, it means orchestrating even larger, more ambitious productions

3. In investing, this is about avoiding the herd and finding the diamonds in the rough

These are not a way to get rich quick, or succeed quick. They’re a way to get rich slow.

There is a different bar for stuff you build vs. stuff you ship

It’s amazing that there’s now a lower barrier to entry with building stuff. I’m literally exhilarated by it. I believe people should build lots of stuff, especially to learn and experiment.

But if you ship everything you build / write / create you’re probably shipping slop. Photographers take 20 photos for every one they use.

So even if you make lots of garbage, don’t ship garbage to consumers / users / customers. (I’ve been responsible for shipping a lot of shit in the past, and I’m constantly trying to do better.)

Build with cleverness and authenticity

Clever ways of building and growing things is awesome. Viral and memeable initiatives are super powerful. Brex mailing their early customers a bottle of champagne was a spectacle, but thoughtful. Ramp’s (and let’s not forget Wendy’s) social media accounts are full of personality.

Build with authenticity. If you’re shipping lighthearted / playful stuff, that’s not meant to be held to a high bar, that’s fine! Intent matters.

My favorite example of this is Pieter Levels (@levelsio). He’s an incredibly prolific builder. He moves on from the stuff that doesn’t work, and makes the stuff that works really good. And it’s very obvious how authentic he is.

And I want to shoehorn in Sarah’s post since it’s so eloquent:

If you want people to buy your stuff, trying being useful

Cold email just doesn't work anymore. Funnily enough, AI-slop-outreach was the final nail in the coffin. Maybe it should have never worked, but it’s certainly dead now.

But people will always respond to generosity, utility, authenticity, and social proof. I’m trying to exercise this by a) building or producing useful / fun things every week, b) investing / giving back to the ecosystem, and c) being an active, contributing member of relevant communities. If this later ends up being effective “business development,” great — but if not, these sorts of activities feed my soul regardless.

Try to create some positivity along the way

I talked to a founder last week, who’s also taking a viral marketing approach. It’s rooted in positivity, not ragebait: they interview immigrants, HONY style. 75k+ followers, and building a business on that foundation that’s already growing really fast. And most importantly: it was clear how ambitious yet humble, viral yet thoughtful their approach was.

I’m not saying everything has to be warm and fuzzy. But there should be a point to virality beyond just eyeballs.

Your reputation is all you have

As AI increasingly takes over more capability, I think humans will retain a monopoly (for a while) over original thinking, kindness, thoughtfulness, and positive-sum games.

Do this over and over, and you establish the kind of brand that people want to surround themselves with.

I think slop in all forms — degrading quality, misinformation, poor signal to noise, bot armies — is the greatest threat to our cognitive health in the coming decade. I hope we can resist the low-calorie, short term hit… and fight to make things that are hard, thoughtful, tasteful, and authentic.

We don’t need more slop. We need more soul.