You have heard the pitch. AI is coming for the SEOs and the web designers, the same way it is supposedly coming for everyone, and soon a business will just type a prompt and get its rankings and its website out of a vending machine. I build with AI every single day, which is exactly why I can tell you the pitch is wrong. AI is a spectacular tool and a genuinely terrible replacement, and the reason is one word: discernment.
Discernment is the core of both jobs. It is also the one thing AI does not have.
First, credit where it is due
I am not here to wave AI off. It is the best leverage our field has ever been handed. It clusters a thousand queries in seconds, drafts a first pass I would have spent an afternoon on, flags technical issues across a site with thousands of URLs, and rewrites metadata I would never hand-type one entry at a time. The repetitive middle of the work, the part that was always more stamina than skill, is getting automated, and that is good news.
So this is not a skeptic complaining that the new tool is scary. It is someone who uses the tool constantly explaining what it still cannot do.
The thing AI does not have
Discernment is knowing which of ten true things is the one that matters here. It is knowing what you do not know. It is the quiet judgment that separates a fact from the right fact. And AI, for all its fluency, has none of it.
The clearest proof is how confidently it gets things wrong. When researchers at Columbia’s Tow Center tested eight AI search tools on 1,600 straightforward source-citation questions, the tools were wrong more than 60 percent of the time, ranging from 37 percent for the best to 94 percent for the worst. The damning part was not the error rate. It was that the tools almost never hesitated, hedged, or declined. They delivered wrong answers with the same polished confidence as right ones. A nervous human intern says “I am not sure, let me check.” The machine does not, because it cannot tell the difference. That gap is discernment, and no amount of fluency fills it.
In SEO, discernment is most of the job
An AI can generate a 300-item technical audit in a minute. It cannot tell you which of those items actually matter for the business, and it will hand you all 300 with equal urgency. It does not know your client’s margins, their most profitable customer, the competitor who just changed strategy, or the real reason traffic dropped in March. It optimizes toward a generic ideal. A good SEO optimizes toward one specific business’s actual goal, which is usually not the same thing at all.
AI will also recommend a textbook best practice that happens to be wrong for this particular site, and it will do it without a flicker of doubt. Someone has to catch that. Someone has to own the call when it goes live, because accountability is not something you can delegate to a model that has no stake in the outcome and no memory of being wrong. Judgment, prioritization, and responsibility are most of what a client is actually paying for. The audit was never the hard part. Knowing what to do with it was, and still is.
In web design, discernment is the whole game
AI can produce something that looks like a website in seconds. Looking like a website and working for the business’s customers are entirely different achievements.
Design is judgment about a human on the other end of the screen. What do they need, what is in their way, what should be removed, and when does something just feel off. AI has never been the frustrated customer trying to book an appointment on a cracked phone in a parking lot, so it cannot feel when a layout fails that person. It generates plausible. It cannot feel wrong. Taste, restraint, accessibility done properly rather than performatively, and the strategic call about what the site is even for, all of that is discernment, and it is the entire difference between a page that looks like a website and a tool that actually earns a business money.
Even Google is factoring in the human
Here is the part that should settle the argument for anyone who still thinks the machine wins on pure output. The algorithm itself is leaning toward the human.
Google added “Experience” to its E-E-A-T quality framework, and its guidance now explicitly rewards content that demonstrates first-hand expertise, the kind that comes from having actually used a product or visited a place. Think about what that means. Search is trying to reward a quality signal a machine with no human experiences cannot honestly produce. The system is not moving away from the human behind the work. It is moving toward them.
The tool is only as good as the hand on it
There is a quieter version of the replacement myth worth killing too: the idea that AI is a fixed amount of smart, and you just turn it on. It is not. The same model hands a novice a confident, generic mess and hands an expert something genuinely useful, because the output is a direct reflection of the person driving it. AI is only as intelligent as the human directing it.
That is the part people skip when they assume the tool levels the playing field. Knowing what to ask, what context to load, what to push back on, and how to spot the moment it has gone confidently off the rails is a skill, not a default setting. Prompting well is closer to knowing how to coach a sharp but very literal junior intern than it is to typing a wish into a box. Hand the same model to someone with real SEO judgment and someone without, and you do not get the same website or the same strategy. You get two very different things, and the entire difference is human.
Multiply, don’t replace
So the real choice was never human or AI. It is whether you put AI at the center or the human at the center.
AI is a multiplier, and a multiplier needs something real to multiply. Point it at genuine expertise, a real point of view, and actual first-hand experience, and it extends all of that further than one person could reach alone. Point it at nothing, ask it to supply the judgment as well as the sentences, and you get confident, well-formatted nothing produced at tremendous scale. I lean on AI hard precisely because it makes a good human faster, not because it makes the human optional. The person brings the discernment. The machine brings the reach. Swap those roles and the whole thing falls apart.
So, is the job going away?
No, but it is changing shape, and the shape is worth understanding. The parts of SEO and web design that were always leverage, the grunt work and the first drafts and the repetitive passes, are getting automated, and I say good riddance. The parts that were always judgment, deciding what matters, what is true, and what genuinely serves this specific person and this specific business, are becoming more valuable, not less.
They have to. The volume of confident, generic, machine-made mediocrity is going straight up, and the only way to stand out in that flood is the exact thing the flood is made of lacking. AI did not kill the SEO or the designer. It just made discernment the entire job, both the judgment to know what is right and the skill to draw something worth using out of the machine in the first place. That, if we are honest, was always the part worth paying for.
Sources
- Columbia Journalism Review, Tow Center for Digital Journalism, AI search has a citation problem (eight tools, 1,600 queries, wrong more than 60 percent of the time, and rarely declining or hedging)
- Google Search Central, Creating helpful, reliable, people-first content (rewards demonstrable first-hand experience and expertise; “Experience” added to E-E-A-T in December 2022)