McCarthyism to the machines: AI’s role in rational thought
Opinion
Much like McCarthy, brands are becoming more accountable because everything they say is now instantly verifiable by a machine that doesn’t care about your media spend or creative treatment, writes Chris Herbert-Lo.
History doesn’t repeat itself, but it often rhymes.”
The McCarthy era isn’t just the tale of a political witch hunt; it’s a media lesson that resonates today.
As Senator Joseph McCarthy (pictured) sought to root out and blacklist communist sympathisers, journalistic norms of the time dictated that reporters quote elected officials without verifying the truth. The press simply amplified McCarthy’s opinions and false claims.
It was a relatively new media channel which eventually broke the whole thing open. A 1954 TV broadcast, ‘See It Now‘, let McCarthy speak for himself, at length, and trusted viewers to notice the problem.
Stripped of the protective gloss of uncritical reporting, the Senator’s own film clips exposed his bullying, shallow tactics and public opinion quickly turned.
The loss of common understanding
Fast-forward seventy years, and where once a compliant press allowed views to circulate as fact, we now have algorithmically curated echo chambers doing the job on an industrial scale. Each of us is now the proud owner of a personalised reality in which our opinions can be continually validated.
In November 2024, the House of Lords Communications and Digital Committee published The Future of News. It warned of the emergence of a “two-tier” news environment in which more engaged and affluent audiences continue to access high-quality, professional journalism, while a growing portion of the population disengages or relies on lower-quality, unverified online content.
The truth is out there. It’s just not evenly distributed.
AI’s ability to break the echo chambers
Perhaps the spookiest parallel between the McCarthy era and today is that when the career-ending ‘See It Now’ broadcast aired, television had just surged into 55% of US homes.
Last month, Ofcom’s annual Media use and attitudes report found 54% of UK adults are now using AI tools following its own recent surge. A rhyme in history that’s clear enough for kids to recite in nursery.
The AI tools in question largely consist of LLMs, which FT reporter John Burn-Murdoch (a top-tier source for news) recently demonstrated are “nudging people away from the most extreme positions and towards more moderate and expert-aligned stances.”
His analysis builds on a body of research suggesting LLMs are converging towards a common truth because, where people might retreat to social echo chambers for validation, AI is expected to be accurate.
The implication is that AI may come to represent a broad centre of collective truth and understanding in a fragmented world, where we can all pick and choose the media around us. Acting as an easy and trusted shortcut for rational thinking in a world where most of our time and attention is captured by more emotive media.
What this means for advertisers
The industry is well-versed in the emotional side of things. Whether it be landing the big AV spots, the contextual billboards, influencing with influencers, or creating with creators.
But much like McCarthy, brands are about to become a lot more accountable, because everything they say is now instantly verifiable by a machine that doesn’t necessarily care about your media spend or creative treatment.
When I ask for the best running shoes, ChatGPT suggests Nike and Adidas, as you’d expect. But it also recommends Brooks (3% share of voice in the running shoes category over the past year) and Hoka (0.4% SOV). Small challenger brands are achieving parity with decades of brand-building in an instant.
It’s the same story when asking about TVs: placing a TCL product (less than 1% share of voice over the last 12 months) on a shortlist alongside Samsung and LG.
Family cars, supermarket pizzas, sunglasses…take a look at your category and see how well it reflects the competitive set whose spending you’ve been tracking for all these years. Then consider that LLMs are only likely to: a) increase in usage, b) become more accurate, and c) link more efficiently to direct purchase.
Whether you’re a legacy brand terrified of challenger incursion or a scrappy upstart sensing opportunity, things are changing.
To succeed means paying attention to how communications build real human sentiment, particularly the kind that gets captured in the environments AI cares about. The example of my running shoes above came from a question asked before the London Marathon. Repeating the experiment after Sabastian Sawe’s record-breaking run receives a response dominated by variants of Adidas Adizero without a Nike shoe in sight (though Hoka is still very visible).
Getting it right requires an understanding of the interplay between opinion in emotive spaces, trusted advice from authoritative voices, and what the brand says about itself across every accessible touchpoint.
The emotional feeds the rational. The rational supports the emotional. And somewhere in the middle, AI is taking notes.
Chris Herbert-Lo is a strategy partner at the7stars. Read his new monthly column for The Media Leader on the first Tuesday of each month.