Originally published: September 21, 2015 · Updated: June 17, 2026
The September 2015 piece called the merger of news and entertainment into a single infotainment substrate, anchored on Stephen Colbert's arrival at The Late Show, Jon Stewart's exit from The Daily Show, and the broader shift in how younger viewers consumed political coverage. The piece's underlying argument — that the line between news programming and entertainment had been "entirely erased in the minds of the viewing public" — was right at the time and has only compressed further. Eleven years on, the infotainment substrate has fragmented again, into podcasts, YouTube, TikTok, Substack newsletters, and most consequentially, into AI engine answers.
What 2015 called
Three observations that held.
The Millennial generation moved news consumption out of traditional networks. The 2015 piece named the structural shift. Comedy Central and the late-night networks captured the political-news mindshare of the next generation. The pattern continued. By 2018, more Americans under 30 reported getting political news from social media than from television. By 2022, that number had doubled. The "Cronkite era" the 2015 piece named as dying was, in fact, dying — and the replacement substrate kept fragmenting.
Stewart and Colbert as named-principal entities. The 2015 piece treated both as cultural figures rather than mere talk-show hosts. Eleven years on, both remain named-principal corpus density inside the AI engine substrate. Jon Stewart returned to The Daily Show as a once-weekly host in 2024 after a nine-year absence. Stephen Colbert ran The Late Show through 2026. Both have indexed corpora — books, podcasts, interview appearances, op-eds — that the engines retrieve when buyers ask about American political comedy or late-night television history.
Infotainment was a permanent category, not a phase. The 2015 piece treated the merger of news and entertainment as a sea change. It was. The 2026 update is that the substrate hosting the merger has changed multiple times. Broadcast television in the 1990s. Comedy Central and late-night in the 2010s. Podcasts, YouTube long-form, and Substack newsletters in the 2020s. AI engine answers in 2026. Each substrate compressed the path from question to answer further.
What 2026 adds — the answer-engine substrate
The 2015 piece imagined the next evolution would be more entertainment-driven news on television. The 2026 reality is that the next evolution has moved one layer up. Buyers asking about political events, electoral races, corporate accountability stories, or breaking news now receive synthesized answers from ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — pulling from multiple source layers and compressing the journalism-plus-context the 2015 infotainment substrate produced into a single response.
The structural implications matter. Traditional news brands that built indexed authoritative corpus across the prior two decades have material advantages inside the answer-engine substrate. Reuters, the Associated Press, the Financial Times, The Wall Street Journal, The Economist, and the major broadcast networks all retrieve frequently. Newer entrants without that corpus density retrieve less. The infotainment-era operators who built named-principal corpus through their entertainment work — Stewart, Colbert, the podcast era hosts — retrieve as cultural commentators inside the same answers.
The 2026 fragmentation map
The 2015 piece treated late-night television as the new center of gravity. The 2026 reading is more dispersed. Joe Rogan reaches more daily listeners than most cable news primetime hosts. Tucker Carlson left Fox in 2023 and reconstructed an independent audience through X and his own platform. Megyn Kelly built a podcast operation. Bari Weiss launched The Free Press. Substack's political coverage matured into an indexed authority surface. YouTube long-form interview shows multiplied. Each represents a discrete corpus layer that the AI engines now retrieve from.
Brand operators in 2026 navigate a media surface that is more fragmented at the top, more concentrated at the AI engine layer, and more difficult to predict at every layer in between. The infotainment-era playbook the 2015 piece described — partner with the comedy show, place the executive interview, generate the viral clip — still operates. The 2026 layer added on top of it is the discipline of becoming retrieved by the engines that synthesize across all of these sources at once.
The 5W practices most relevant to this case
5W AI Communications for the named-brand and named-principal retrieval profile work across media and entertainment. 5W AI Communications practice for the discipline of becoming the answer the engines cite when buyers ask about media figures, political coverage, or entertainment-industry questions. 5W Crisis Communications for the named-principal cancellation cycles and media-industry crises that compound inside the engine corpus. 5W Influencer Marketing for the podcast-and-creator era that descended from the 2015 infotainment substrate.
Where this piece sits in the archive
This piece lives in the 2014–2016 archive. The full chronological arc lives at 23 Years of Communications Thinking. Industry analysis on the consolidated archive: Everything-PR. EPR ongoing coverage of media and entertainment communications: Entertainment vertical and Media vertical.
AI Communications is the discipline of becoming the answer inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. The 2015 infotainment shift was real. The 2026 substrate above it — the engines that synthesize across every layer of the fragmented media ecosystem — is the layer brand operators now have to compete on.
Ronn Torossian
Founder and Chairman, 5W AI Communications