Internet companies don't owe anyone free traffic

It's incredible people still haven't realized this and hedged against it

I was reading this Bloomberg piece over the holidays on AI taking traffic from food bloggers and, incredibly, it’s framed like a tragedy. Perhaps on some level it is, but it’s not exactly a shock. Plucky food bloggers, years into their craft are blindsided by AI “recipe slop” that supposedly burns cakes and wrecks Thanksgiving. The real story is simpler and less sentimental: an entire cottage industry built on ranking first for “how long to cook a turkey” is finally colliding with a tool that can answer “how long to cook a turkey” instantly. And the reality is for basic, well-trodden dishes, AI works fine. If you need stuffing, mashed potatoes, or green bean casserole, you are not in the market for a 2,000-word essay, five autoplay videos, and three banner ads before you get to “preheat oven.”​

These bloggers didn’t just get unlucky, and what is happening shouldn’t shock anyone. They optimized their entire business around Google’s incentives. Longer posts, keyword stuffing, life-story intros, endless ads – all in service of rankings and monetization, not user delight. Now the same platforms that rewarded that behavior are doing what platforms always do: changing rules when it suits them. AI Overviews pull the useful bits, strip out the clutter, and keep the user on Google. Of course that cannibalizes their traffic. That was the logical endgame the moment you made “free traffic from another company” your business model.​ Similar story with Facebook, LinkedIn or anywhere else you do not host your own work.

No one owns “how to roast a chicken”

A lot of the quotes in the story carry this vibe: “we tested these recipes, AI didn’t, so it’s unsafe and unfair.” That might land for niche, technical, or culturally specific cooking where expertise actually matters. If you’re writing deeply researched guides on regional Chinese ingredients like The Woks of Life, it absolutely sucks to watch years of work flattened into a generic paragraph that now outranks you. But for the vast majority of mainstream home cooking, we’re not talking about proprietary IP. It’s chicken, pasta, brownies, turkey. These are commodity tasks.​ And anyway, the internet already has copycats of everything.

The uncomfortable truth is a lot of food blogging success was simply timing plus arbitrage. You got there early, wrote decent recipes, then enjoyed years of free traffic because consumers had no better options than “click the top blue link.” Hard to ever call that much of a moat, it was simply being early. The internet finally behaves closer to how users actually think: “tell me quickly what to do, don’t make me scroll your memoir to find the ingredient list.” The idea that anyone is entitled to ongoing attention for generic “how to make lasagna” knowledge was always tentative.​

Platforms don’t love you back, no one ‘owes’ you traffic for existing

The most revealing parts of the Bloomberg story aren’t about recipes, they’re about power. AI Overviews remix ingredients from Inspired Taste with instructions from competitors, then show the Frankenstein result above the original source, even on brand-name searches. Pinterest’s algorithm pushes AI-generated food images and cloned recipes ahead of the human originals, and users are getting tricked by impossible dishes with broken instructions. Facebook and random content farms scrape photos, lightly rewrite text with AI, and spin up fake sites that pass plagiarism checks while siphoning ad revenue.​

If that sounds familiar, it should. This is the same playbook that hammered publishers, news sites, listicle farms, and YouTube. Build on rented land, get hooked on platform distribution, watch as the landlord one day says “actually, we’re keeping more of this for ourselves.” No one owes you traffic just because you’ve been around for a few years. “I built my entire livelihood on an opaque ranking system run by trillion-dollar companies” was never going to end well. This is also why I am constantly in praise of email, which no single company controls. I agree the platforms should rank and serve high quality content and link to independent publishers, by the way. But they altered the deal, and here we are. If there ever was any sort of internet social contract, big tech tore that up ages ago.

AI slop is real too, but that’s not the point

There is real slop, and I’m critical of it all the time. There are AI recipes that tell you to cook a 6-inch Christmas cake for 3–4 hours at 350°F, which is a nice way to manufacture charcoal. There are AI tamale photos with sauce poured over the husks and steaming them flat like envelopes, which any human with even passing familiarity would flag as wrong. There are cloned food blogs, uncanny family photos, and synthetic versions of creators’ kids staring out from sites they never built. That’s creepy and gross. It’s also exactly what black-hat operators do whenever there’s an arbitrage window: identify a high-traffic category, spin infinite content, capture just enough clicks and ad impressions to make the math work.​ Recently people have been dumb enough to brag about this publicly, but usually you never know the people doing this for obvious reasons.

Still, the existence of slop doesn’t magically justify the old status quo. The argument can’t be “slop is bad, therefore endless-scroll ad farms are good.” Users hated the clutter long before AI. That’s partly why AI answers are winning. For the average person trying to get dinner on the table, the hierarchy of needs is: 1) don’t poison my family, 2) don’t waste my time, 3) don’t shove intrusive display ads in my face. On those goals, a halfway competent AI beats a SEO-maxed blog 9 times out of 10. The right critique isn’t “AI should go away” but maybe “platforms should stop laundering scraped work into their own products without meaningful attribution, traffic, or payment.” And for simple recipes, I’m sorry but I don’t see how anyone wants to go back to a site with a terrible experience and difficult to find ingredients list.

Where the real opportunity is now

If search is over for generic recipes, then so is the illusion that “I have a blog” is a strategy. The new game is the old game and how marketing used to be: build a brand, own distribution, and give people a reason to care about you beyond “I rank first for how to cook a turkey.” The food creators who survive this shift will look less like anonymous SEO ghosts and more like Gordon Ramsay: personality-forward, story-driven, multi-channel, and not dependent on one algorithm’s mood. His YouTube channel is amazing by the way, has >20M subscribers and I still watch his videos all the time even though I could ask AI for a dish.

Anyway, this means:​

• Owning your audience via email, direct subscriptions and a following of fans who will promote your work.

• Leaning into video, live formats, and social where your face, voice, and values are the moat, not just “best pumpkin pie recipe.”​

• Creating depth and specificity AI can’t easily flatten: regional cuisines, technique-driven education, behind-the-scenes content, commentary, and cultural context.​

There’s still plenty of room for human-led food media, but it looks more like “this is my creative world, come hang out with us” than “here is my 8th variation on sheet pan chicken for Tuesday SEO.” People will absolutely still follow cooks they trust. They just won’t find them by typing “how long to cook turkey” and clicking the top result and getting a page filled with ads and spam.

The broader lesson: don’t be a feature of someone else’s product

Zoom out and this goes way beyond recipes. Anywhere the value is a commodity instruction set – how to reset a router, write a basic cover letter, tie a tie, choose a mattress – AI is going to win against the long tail of undifferentiated “content.” Anyone whose business model is “I exist because I ranked early for a generic problem” is on borrowed time. If your work can be reduced to a paragraph inside someone else’s interface, eventually it will be.​

The Bloomberg story reads like an obituary for a certain era of the web. That era rewarded people not for being the most interesting or original, but for being early, prolific, and good at gaming the machine. Now a bigger machine showed up. The lesson is not “technology is evil” or “AI must be stopped.” It’s: never mistake borrowed attention for durable equity. Don’t build your house on stilts next to the sea, then act surprised when a hurricane comes.