BusinessWeek, as the publication once spelled its name, was decades ago a must-read magazine, chock full of intelligent analysis. While The Wall Street Journal focused on reporting facts, BusinessWeek helped readers make sense of them and put events in perspective. I devoured the magazine cover to cover on Friday nights when it landed in my USPS mailbox.
BusinessWeek was especially strong on technology, and not the cheerleading masquerading as journalism today. One prescient story that has stayed with me some thirty years later examined Silicon Valley’s “almost is good enough” mindset. Tech culture, the piece explained, prized being first over being right, which explained why so many products shipped riddled with bugs. These were not considered failures but works in progress to be fixed on the fly.
In consumer technology, “almost is good enough” is more easily tolerated – and these days even expected. When an app glitches, users complain and wait for the next update. But when that same mindset migrates into institutions where reliability, judgment, and trust are the product, the consequences are more serious.
Journalism is now one of those institutions.
The Washington Post’s AI Debacle
The Washington Post offers a telling example of what happens when product logic overtakes editorial judgment.
As reported by Semafor, the Post recently began publishing AI-generated podcasts even after internal testing showed the technology introduced errors and bias into the paper’s reporting. The feature, called “Your Personal Podcast,” struggled to meet the Post’s standards from the start.
Across three rounds of internal testing, between 68 percent and 84 percent of scripts were rated as failures. Errors ranged from mispronunciations to misattributed or fabricated quotes. In some cases, the tool inserted commentary by interpreting a source’s remarks as the paper’s position on an issue. Internal reviewers concluded that small prompt changes were unlikely to meaningfully improve results without introducing additional risk.
Despite those findings, the podcasts were launched anyway.
The initiative was overseen by the Post’s chief technology officer, Vineet Khosla, and its head of product, Bailey Kattleman. Khosla previously worked as a senior engineering manager at Uber and holds a master’s degree in artificial intelligence from the University of Georgia. Kattleman has spent years in product development at the Post and previously held similar roles at NPR and Major League Baseball. She holds a degree in English and media studies. Neither has worked as a journalist.
Those backgrounds are not disqualifying. But they help explain how newsroom standards could be overridden by product imperatives, and why practices that would be unthinkable for human reporters were treated as acceptable for product software.
If legendary editor Ben Bradlee of Watergate fame were alive and still running the Post, the project would never have seen the light of day even with a 95 percent accuracy rate. Under Bradlee, accuracy was not a metric to be optimized. It was a condition of employment. A reporter who was wrong five percent of the time would not have lasted long at a major publication.
What changed was not journalism’s mission. It was who was empowered to decide how that mission would be carried out.
The Post, of course, is owned by Jeff Bezos. Amazon’s success was built on relentless experimentation, optimization, and automation. Those instincts transformed retail and logistics. Applied to journalism, they are far less benign.
Journalism is not a workflow. It depends on discipline, judgment, and interpretation. Treat it like software, and “almost right” becomes acceptable, leaving the Post unable to justify its already discounted subscription price.
The New York Times’ Platform Mindset
Technology has become increasingly critical to the New York Times’ survival. The Times has embraced it not merely as a tool in service of journalism, but as a parallel business that now underwrites it. Games, Cooking, newsletters, audio, apps, and bundles have become as central to the Times’ success as reporting itself.
The Times now describes itself less as a newspaper than as a “subscription platform,” with journalism as the anchor tenant but no longer the sole draw. When reporting fails to break through, Wordle still refreshes. That diversification has brought revenue and resilience, but it has also reshaped the institution’s culture.
That shift surfaced just before the presidential election, when technologists at the Times walked off the job during a contract dispute. Journalists historically understood their work as a public trust, especially at moments of national consequence. The technologists who walked out appeared to view their roles as platform labor, detached from the stakes of the moment.
The distinction matters. It reflects a newsroom no longer unified around a single mission.
The Wall Street Journal’s Engagement Journalism
The Wall Street Journal was once America’s preeminent newspaper for accuracy, fairness, and depth for its long-form reporting. Under editor Emma Tucker, an import from the UK, the publication has increasingly adopted the rhythms of a scoop-driven tip sheet, with “Exclusive” banners blaring across its homepage and engagement metrics shaping editorial priorities.
Where Journal reporters once had to do considerable work to earn space in the paper, they are now expected to produce scoops at a pace that would once have been unthinkable. Not surprisingly, veteran reporters and editors, including Pulitzer Prize winners, have been leaving for publications that were once regarded as a step down.
A recent Journal story about corporations hiring in-house “storytellers” to promote their brands illustrated the shift. The piece offered no case studies and little reporting on whether such efforts actually work. In years past, it would have been sent back for more reporting. Instead, it generated social media buzz, which increasingly appears to be the point.
That is not a failure of individual reporters. It is the result of editorial priorities shaped by engagement logic rather than superior journalistic judgment and values.
Substack’s Blurring Boundaries
For a time, Substack appeared to offer something different. The platform attracted respected journalists and offered independence from institutional pressure.
But Substack is not a charity. It is backed by powerful technology investors, including Andreessen Horowitz, and like most venture-funded platforms it faces pressure to scale, monetize, and eventually provide liquidity to its backers.
That pressure is increasingly visible. Substack has begun courting corporate brands and encouraging sponsored content as a growth strategy. This year, brands including Madewell, American Eagle, Billboard, Rare Beauty, The RealReal, Bandcamp, and Tory Burch have launched newsletters on the platform. The company is also encouraging founders to publish newsletters that blur the line between personal voice and brand promotion.
The logic behind that strategy is revealing.
“We’re always looking for bringing the most important, influential voices to the platform,” Christina Loff, Substack’s head of lifestyle partnerships, told Marketing Brew.
The quote is striking for what it does not distinguish. In Substack’s framing, a journalist doing original reporting and a brand founder promoting a product are both simply “voices.” Influence, not standards, is the unifying value.
Not surprisingly, Loff’s LinkedIn bio notes that she has spent most of her career in PR and content marketing. One might have expected a platform positioning itself as a home for serious writers and reporters to place an accomplished journalist in charge of recruiting and curating prominent voices.
When companies can publish marketing material on the same platform that hosts serious reporting, the boundary between journalism and promotion erodes, cheapening the perceived value of journalism itself.
R.I.P. BusinessWeek
BusinessWeek was eventually acquired by Bloomberg. The publication still produces some compelling reporting, but I can’t recall reading a story there in recent years that left me confident I’d still remember it three decades from now — even if I’m still around.
Artificial intelligence is not remotely capable of producing the kind of analytical journalism BusinessWeek once did. And based on my own experience using it, I remain skeptical that it ever will. The technology can be impressive, even dazzling at times. But it repeatedly lacks the common sense and human empathy that great journalism requires.
That limitation isn’t so much a failure of technology as a reminder of what journalism once valued. Accuracy over speed. Judgment over output. And the assumption that readers would rather wait for something worth remembering than be fed a constant stream of what’s merely good enough.