Artificial intelligence in 2026 is a normal technology - not a job-apocalypse, not a superintelligence - and its disruptions are no more fantastic than those of trains or cars. That's the core of a recent analysis published June 29, 2026, arguing that journalism's job is not to make AI more accurate, and that only public policy can sustain a "learning society" where knowledge creation and community thrive. The argument has direct stakes for writers, whose work feeds the very systems that now threaten their livelihoods.
AI is a normal technology
The analysis pushes back on years of hype. "AI causes disruptions much like any other innovation. Its impacts, while real, are not fantastic or mysterious," it states. The familiar drives of human nature and capitalism remain unchanged. AI is simply software, here like computers and cars, with good and bad effects for users and non-users alike. Decluttering the rhetoric, the argument goes, is essential to have plain conversations about AI's role in society.
The learning society and the knowledge paradox
Economists Joseph E. Stiglitz and Bruce C. Greenwald's concept of a "learning society" - where widespread knowledge raises living standards, fuels innovation, and enriches art - doesn't happen by accident. It requires public policies, institutions, and norms that support lifelong learning. But markets for information are inherently inefficient because of the Grossman-Stiglitz Paradox: the ease with which information spreads often prevents its producers from recouping costs. This leads to paywalls, patents, and underinvestment in knowledge creation.
AI, much like libraries and search engines before it, aims to be an ultimate spillover technology, collecting what has been thought and said and delivering it efficiently. Yet in 2026, it presents serious problems for a learning society. Users overestimate its capabilities. It encourages cognitive offloading in schools and workplaces, boosting short-term productivity at the expense of longer-term learning. AI harvests not just content but the audiences that fund further knowledge creation. Content pollution has eroded trust: teachers suspect students of cheating, freelancers are doubted as AI fakes, and slush piles are dismissed as slop. When AI eliminates jobs, it also eliminates the ability of workers to learn by doing. In short, the evidence suggests AI draws from the pool of public knowledge without restocking it.
Why journalism shouldn't prioritize AI accuracy
Gina Chua recently argued that "our energy should be spent on making them more accurate, not just complaining that they aren't." The analysis disagrees, particularly from a news industry perspective. Making AI a better source of current-events information turns it into a substitute good for news providers who funded the original reporting. Even if AI companies paid licensing fees, many news outlets may rationally avoid participating - the transaction costs for smaller publishers are daunting, and the net revenue is uncertain.
Moreover, a world without publishers and journalists loses more than information production. News organizations serve as collectors of what Alexis de Tocqueville called the "wandering spirits" of democracy, building community alongside knowledge. AI, by contrast, is inherently anti-community. Twenty years of Facebook and Twitter show that tech development does not bend toward accuracy - it trends toward slop for the sake of market share. The history of social media and journalism is a graveyard of sunk costs, with many journalists still stuck on platforms like X due to status-quo bias, even as public opinion on AI sours.
Public policy, not market forces, will build the learning society
The analysis points to a modest but telling example: high school students at a journalism camp resented their time on social media and knew it wasn't trustworthy. One student became an avid New York Times reader through a school subscription program - essentially a public subsidy that bypassed the paywall. That intervention corrects the Grossman-Stiglitz Paradox, aligning funding with the sharing of accurate information.
Whether AI becomes a more accurate source or news organizations survive to serve democracy's wandering spirits is not a given. It's not clear both futures can coexist. Public policy - taxes, subsidies, institutional support - will likely need to play a role. The learning society is something we have to want to build.
Why this matters for writers
For writers, the normalization of AI means the technology is neither a savior nor an existential threat, but a tool that demands deliberate choices. The argument that newsrooms shouldn't rush to make AI more accurate underscores a strategic reality: feeding the machine may accelerate your own obsolescence. Writers grappling with these shifts can explore resources on AI for Writers to better understand the tools affecting their field. Meanwhile, the erosion of trust in creative communities - from slush piles suspected of AI slop to freelancers doubted as fakes - is a growing concern, and guidance like AI for Creatives can help navigate those challenges. Ultimately, the analysis suggests that protecting the conditions for learning, community, and accurate information will require writers to advocate for public policies that sustain the knowledge ecosystem, not just adapt to its disruption.
Your membership also unlocks:
