Originally published on Jax5d in December 2025 as part 3 of Jacquie Baker’s five-part series on the first 1,000 days of mainstream AI. Adapted and expanded for Probably Genius.
Staying out of AI does not keep AI from talking about you. Assistants answer questions about businesses either way, assembling their descriptions from whatever the record holds: directories, old profiles, reviews, other people’s pages. Opting out only removes your version from that record. The real choice in front of every owner is authorship or absence, and absence is still a description. It is just one you didn’t write.
Some owners have a plan for the AI era, and the plan is to sit it out. Don’t feed the machines. Don’t put more of yourself online. Keep your head down, protect your privacy, and wait to see how all this shakes out.
Let’s give that instinct its due, because it is not stupid. It usually belongs to careful people: professionals who have watched platforms chew up their industries, who dislike performing for algorithms, who believe good work should speak for itself. The instinct is protecting something real.
It just rests on one assumption, and the assumption is wrong. Hiding assumes that if you stay out of the conversation, there is no conversation. Our co-founder Jacquie Baker took this argument apart in a December 2025 essay, and at Probably Genius we now watch its mechanics play out in diagnostic work every month. This guide covers what the machines do with your silence, what the leftover record actually says, and what taking authorship looks like for an owner who genuinely values privacy.
The Plan to Wait It Out
The waiting strategy has a respectable pedigree. Plenty of good businesses sat out early social media and lost nothing that mattered. The owners who skipped the metaverse look like prophets. Sitting out a technology wave has worked before, so why not this one?
Because this wave is different in one specific, structural way: the platforms you sat out before only talked about you if you showed up. Facebook did not write a profile for your firm and answer questions about it in your absence. AI assistants do exactly that.
When a buyer asks an assistant about your business, or your category, or the problem you solve, the machine does not check whether you opted in. It answers. The only thing your abstention changed is what material the answer gets built from.
Jacquie’s essay put the stakes in one line that has been quoted back to us by owners ever since: “The real danger isn’t being seen; it’s not being seen by the right people.” The hiding strategy does not manage the danger. It hands the microphone to whoever else has ever written anything about you.
You Are Already in the Answers
Here is the machinery your silence is up against.
First, the machines have been building files on businesses for years without asking anyone. Google describes its Knowledge Graph as a database of billions of facts assembled automatically from open-web sources, with entries generated whenever enough information exists. Read that sentence from the abstainer’s chair: the file exists whether or not you contribute to it. Your participation was never the trigger. Other people’s information was.
Second, the assistants sitting on top of that record would usually rather answer than abstain. When the Tow Center for Digital Journalism tested eight AI search tools in March 2025 on a basic factual task, the tools collectively got more than 60 percent of queries wrong, and the study’s most striking finding was the posture: the engines delivered wrong answers with confidence instead of declining to answer. An assistant asked about your business does not tend to say “this firm has chosen not to engage with AI.” It says something. Fluently. With the same straight face it uses for facts.
Put those two together and the hiding strategy inverts itself. Silence doesn’t make AI say nothing about you. It makes AI say whatever everyone else left lying around.
What the Leftovers Say About You
So what, exactly, has everyone else left lying around? For a typical established firm that “stays off the internet,” the record the machines can reach looks like this: a directory listing from your founding year with the old suite number. A profile on a platform you abandoned. A handful of reviews, the angriest one written the most vividly. A former employee’s page still listing your firm. A competitor’s comparison table where you appear as the expensive option, described in their words.
That pile has three properties worth noticing. It is stale, because you stopped tending it years ago. It is scattered, so the machine has to guess at which fragments even belong to the same business. And none of it was written to represent you, so every line of it serves someone else’s purpose or no purpose at all.
We have a name for what that pile produces: the Training Data Echo, the pattern where AI confidently describes the business you used to be, because old evidence outvotes a current story nobody put on the record. Hiding does not protect you from the echo. Hiding is the echo’s supply chain. Every year you abstain, the stale record ages one more year while remaining the freshest thing the machine can find about you.
And the audience for this pile is not hypothetical. It is your next client, holding the private consultation your traffic stats can’t see, taking the machine’s description as a briefing. The buyer does not know you opted out. They just hear an answer with your competitors described in their own chosen words and you described in leftovers, if you appear at all.
Authorship, Not Absence
The way out is not louder marketing, and nobody is asking you to become a content creator. The way out is authorship: making sure the version of your business the machines read is the accurate one, written by the person who actually knows.
The principle carrying this entire series is the one Jacquie published two weeks before the hiding essay: AI can’t cite what you won’t claim. Hiding is the contrapositive lived out. Refuse to claim your work, and the machines quote around you: the directory, the angry review, the competitor’s table. Authorship reverses the flow at its source.
In practice, authorship is three unglamorous moves. Establish one coherent identity, so every surface the machine checks agrees on who you are, where you are, and what you do; we wrote a plain-English guide to how AI decides your business is real. Claim your positions and your method under your own name, on pages built to be quoted. And put your proof on the record with sources, so a machine checking your claims finds confirmation instead of silence.
Notice what happens to the effort curve once the record is right. Jacquie’s essay described the payoff from her own practice: “my ideal clients find me without me having to explain myself.” And her line about why accuracy beats spin holds the whole strategy together: “Truth holds its shape without you having to hold it.” A record built from your real work does not need defending, updating, or performing. It just needs to exist.
She also wrote about what the visible version of this buys, in the most concrete terms she had: “When my husband and I moved our three kids from Melbourne to the USA for a year of snowboarding, that wasn’t a mindset—it was financial independence.” Then the line under it: “Freedom isn’t a concept you believe in. It’s choices you can actually make.” A business the machines can find, describe accurately, and recommend is a business that keeps working while you live your life. That is what was actually at stake in the hiding decision all along.
Privacy for Your Life, Authorship for Your Work
Now the part the privacy-minded reader has been waiting for, because it matters and it is honest: none of this requires giving yourself up.
Your family, your politics, your weekends, your face on camera: keep all of it private, permanently, with our blessing. The machines do not need your life. They need your work: what you do, who you serve, what you charge, what you know that your competitors don’t, and proof that it is true. Authorship of the business record and exposure of the person are different things, and the entire strategy runs on the first one.
The same boundary applies from the other direction, and we hold it as policy. Getting the record right never means burying the parts you dislike. The honest version of this work updates what the machines believe by giving them better evidence: current facts, claimed positions, verifiable proof. Suppression has no place in it, and we wrote that into our AI Integrity Standard because a recommendation built on anything else eventually collapses on the business it was supposed to help. Accuracy is the entire job.
So the question was never “engage with AI or protect yourself.” Protecting yourself is the reason to engage. Somewhere in the record right now is a version of your business assembled from leftovers, and it is being read aloud to your buyers. You are probably better than your leftovers. Authorship is how the machines find that out.
Want to Learn More?
Probably Genius was built by Jacquie Baker and Christopher Shaw, who spent more than 40 combined years translating what makes an expert the best in the room for the audiences that decide who gets chosen. Jacquie has been publishing the thinking behind this work on her Substack, Jax5d, since 2025; this essay adapts the third part of her series on what the first 1,000 days of mainstream AI changed for experts and the businesses built on their judgment.