Chapter 3: Where We Are and Why Now

Picture two leaders you probably know.

If you lead people who depend on your judgment about information tools, you have almost certainly met both temperaments—maybe on different weeks in the same month.

The first is overwhelmed: every week brings a new model headline, a new policy panic, a new message from someone who is counting on you to get this right. They are not lazy—they are carrying a full plate in a foggy weather system. Their temptation is avoidance: “If I do not engage, I cannot get it wrong.”

The second is overconfident: they found a workflow that works for them—fast decks, fast posts, fast replies. Their temptation is absorption without calibration: mistaking throughput for wisdom.

Both postures are understandable. Both are dangerous.

Avoidance cedes the field: your people will still use AI; they will just do it without your relational or organizational scaffolding. Ungoverned confidence ships mistakes at scale—often fluent, rarely humble.

Chapter two named the kind of challenge this is—adaptive, moral, communal, and technical in execution. This chapter’s job is to situate us in time and infrastructure: not to scare you into urgency, but to name why “later” is not a neutral strategy for people whose ideas must be found, trusted, and carried.

A simple time marker

OpenAI’s ChatGPT reached broad public use after its November 2022 release—an ordinary product launch that became an extraordinary social fact. However you date the “start,” generative assistance moved from lab curiosity to default office companion with a speed few institutions had time to digest. There is no elder generation with thirty years of lived practice in this stack. That is not romantic; it is descriptive. We are early.

Discoverability shifted under your feet (again)

For decades, movement leaders have had a complicated relationship with search-engine optimization—sometimes rightly. Metrics can reward thinness; platforms can extract. Much of that critique still stands—and it helps explain why many thoughtful leaders deprioritized SEO: not because discoverability was unimportant, but because the loudest version of the game often rewarded what you did not want to become.

Underneath the slogans, the honest description is simpler: discoverability is infrastructural—crawl paths, intent-shaped phrasing, internal links, entity consistency, and the slow work of looking like a serious publisher—not a moral reward for being first to have the insight. That was true before generative search; it is more true when answers are assembled from many sources at once.

What changed with generative interfaces is the victory condition. Even when you rank well in classic search, users may get an on-page synthesized answer that satisfies their intent without a click. Industry analyses (vendor-dependent, geography-dependent) point in one durable direction: attention is harder to earn and easier to intercept than it was in the ten-blue-links era. Treat the exact click curves as directional: they move with UI experiments, query mix, and geography.

That is why a parallel research agenda—often called GEO (Generative Engine Optimization)—exists. Aggarwal et al.’s foundational paper (arXiv:2311.09735; presented at KDD 2024) introduced benchmark tasks for visibility in generative engines and reported material but bounded lifts from structured interventions—with domain dependence explicitly noted. Follow-on work and vendor benchmarks vary in quality; some will be serious, some will be content marketing dressed as science.

The practical takeaway for you is not “learn every acronym.” It is this:

Being right and being quotable by a machine in a helpful way are related—but not identical—to being findable.

Google’s Search Quality Rater Guidelines remain the clean primary reference for Experience, Expertise, Authoritativeness, and Trust (E-E-A-T)—not a transparent “score” you can game like a title tag, but a north star for defensible substance: clear authorship, transparent sourcing, editorial standards, pages that reduce harm. Google’s own documentation stresses trust as weighted heavily and repeats that E-E-A-T is not a single ranking factor—quality raters align automated judgments; they do not hand-dial your URL up or down. The practical work stays boring: bylines, sourcing, harm reduction, and pages a careful reader would respect. That orientation helps both classic search and the extractive tendencies of models that prefer crisp, well-segmented passages.

Networks relocate the problem—they do not magically delete it

A curated network of leaders can help co-citation, cross-linking, and mutual public recognition—slow trust that resembles your actual intellectual graph more than a million templated FAQs. Done poorly, a network multiplies noise. Done well, it increases distinct coverage and human proof.

If you lead a congregation rather than a distributed network, the same weather still hits you: people search, ask assistants, and forward links before they ever walk your hallway. If you lead a network or nonprofit, you may feel the graph layer more acutely. In either case, I keep landing on the same underlying map: platforms worth building are not only for a lone hero URL—they are for an organization and the leaders and partners orbiting it, because the linked graph (who points to whom, who is quoted alongside whom) is increasingly part of how durable expertise gets surfaced. That is not a promise that “community SEO” replaces depth. It is a warning that depth without legible, connected public presence loses more often than it should.

Networks still need governance: a hundred near-identical explainer pages can cannibalize the very authority you thought you were scaling. I will develop that later as scenius—shared, networked intelligence inside a real field of practice—and bounded collaboration. For now, hold the modest claim: infrastructure matters. Solo heroic URLs were never enough for every leader; they are less sufficient when answers are assembled upstream of your homepage.

A word on hype numbers

Internal strategy models sometimes illustrate reach with large multipliers (including illustrative “network math” scenarios). Unless and until those ratios are published with transparent methodology (definitions of reach, deduplication across overlapping audiences, time window, channel mix), treat them as scenarios, not laws of nature—even when they are offered in good faith as teaching devices. Public credibility is built on what you can show, not what sounds inevitable.

Likewise, beware rhetorical compression like “70–80% left on the table” for email or social unless tied to a named study’s definitions. Safer language: rented audiences, algorithmic intermediation, and measuring what you own (newsletters, communities, events).

Why this matters now (without manufacturing panic)

• Your board, senior staff, students, and donors are already forming opinions in the same low-trust weather chapter one mapped: in Pew’s September 2025 survey of U.S. adults, high stated importance of knowing whether content is AI- or human-made sat alongside low self-reported confidence in being able to tell—the “importance outrunning self-efficacy” gap.

• Your discoverability stack is aging even if your theology is not. If your work is invisible to the ways people search and ask, you will feel falsely “behind” on merit.

• The window for thoughtful policy and practice is still open—early enough to build norms, late enough that delay has costs. That is kairos in the unglamorous sense: not hype about inevitability, but a season where what you institutionalize next will stick.

Where we go next

We have sketched the crisis (chapter one), the kind of problem this is (chapter two), and the landscape shift that makes “ignore it” costly and “hype it” reckless.

Which raises an honest reader question: why listen to any particular voice in this noise—including mine?

That is not a cynical question. It is the right question. The next chapter is about guides: what trustworthy accompaniment looks like when nobody has forty years of practice with generative assistants—and how to spot the difference between a salesperson and a fellow traveler.

---

Reflection questions

1. Are you closer right now to overwhelmed avoidance or uncalibrated confidence? What would move you one step toward the middle?

2. What are five queries your actual audience uses when they look for help you provide? When did you last check what appears in search results—and what a major AI assistant says—when someone asks in plain language?

3. What is one “E-E-A-T” upgrade you could make this month that is boring, visible, and true (byline, sourcing, fixing a stale bio)?