AI is now the most heavily recommended tool in marketing, but much of the advice about how to use it is often questionable. That is the central argument in a new MarTech commentary, which urges marketers to apply the same skepticism to AI thought leadership that they learned during the early years of email automation.
Email marketers have seen this movie before
The piece draws a parallel between today’s AI hype and the formative period of email automation, when marketers had few rules, little data, and no prebuilt sequences. Teams experimented, shared results at conferences and in newsletters, and gradually developed best practices that scaled across industries.
As email matured, it became easier to separate genuine experts from people who simply repeated someone else’s claims. The author argues that AI is now in the same moment: since ChatGPT put AI in front of a mass audience, a flood of quick tips has followed, but far fewer creators are showing evidence that they know what actually works.
Three questions to stress-test any AI claim
The commentary offers a simple evaluation framework:
- Does the advice wildly diverge from consensus? A contrarian idea is not automatically wrong, but be wary if the author cannot explain the departure or show how others could replicate the result.
- Does the expert show their work? Results without process are hard to copy. The author’s own attempt to build a 15-agent system stretched well beyond an initial three-week timeline, reinforcing that the struggle and failures matter as much as the outcome.
- Understand the problem before implementing the solution. Start with “What problem do we need to solve?”, not “Let’s build it.” Ask whether the idea makes sense for your business, whether it is realistic, and whether the outcome is worth the work.
Why a healthy filter matters
AI runs on a token economy. Every query, agent call, and model experiment spends money and capacity. Following unproven advice can revive the “silver bullet” technology mirage marketers have chased for years, wasting budget on tactics that do not hold up.
The commentary anchors the point in an older email case: in 2010, Charles Nicholls of SeeWhy shared data showing the best time to send a cart reminder was an hour after the event, giving marketers the business case to demand data faster. The insight was not a universal rule for every brand, but it was a valid starting point for testing. AI best practices deserve the same treatment: use them as hypotheses, not guarantees.
What to watch for in a thought leader
Beyond the three checks, the author suggests looking at a creator’s background, work experience, publications, brands, and how much they give back through podcasts, webinars, speaking, and mentoring. True thought leaders, the piece concludes, are not only smart; they share what they know to lift others.
For marketing teams, the takeaway is straightforward: read AI advice skeptically, demand process rather than just outcomes, and tie every tactic back to a real business problem before you spend a single token.
Source: MarTech
