AI is killing All About Berlin
AI-generated answers in Google Search and other tools are increasingly replacing clicks to original websites, threatening the viability of small, human‑written resources like the immigration and life guide “All About Berlin.” Commenters argue that search engines and LLMs are effectively free‑riding on years of unpaid or ad‑funded work, accelerating the collapse of an already enshittified, SEO‑driven web and disincentivizing new high‑quality content. Proposed responses range from paywalls, collective licensing and regulation to political activism and a shift toward local or closed communities, alongside a resigned recognition that many ad‑supported business models built on Google traffic may simply not survive.
Impact on “All About Berlin” and Similar Sites
- Many see the site as a high‑quality, human‑curated resource that was crucial for expats.
- Traffic and revenue are collapsing as Google’s AI Overview answers user queries directly, bypassing the site.
- Commenters generalize: ad‑supported, information‑only niche sites that depend on search referrals are broadly at risk.
Role of Google and AI Overviews
- Strong view that this continues Google’s long trend from “links” to “answers,” now accelerated by LLMs.
- Some frame it as monopoly abuse and “enshittification”: Google first depended on websites, then intermediated and now cannibalizes them.
- Others argue it’s “creative destruction”: Google doesn’t owe anyone traffic; users simply prefer faster, cleaner summaries over SEO‑slop pages.
Data Use, Ethics, and Copyright
- Deep resentment that LLMs were trained on public content without consent, payment, or attribution.
- Comparisons to past copyright enforcement: individuals punished harshly for minor piracy while large AI firms face no consequences for mass scraping.
- Counter‑view: much online work was always unpaid (HN, Reddit, OSS); copying and summarizing information has long been considered acceptable.
Future of Content, Incentives, and Paywalls
- Fear of a “dead internet”: if AI slurps up and replaces human creators, new high‑quality content dries up; models then degrade by training on their own slop.
- Expectation of more paywalls, walled gardens, apps, and private communities; open, well‑researched guides may become books, services, or paid newsletters.
- Some argue that only content locked and licensed will remain economically feasible; others think hobbyist and forum content will persist but is less structured.
Broader Social and Economic Concerns
- Anxiety about AI accelerating job loss, hollowing out the middle class, and reinforcing corporate power and surveillance.
- Analogies to enclosure of the commons, industrialization, and “too big to fail” tech oligopolies.
- Minority view: AI is powerful but just another wave like PCs, the web, or mobile; adapt and use it as leverage.
Proposed Responses and Open Questions
- Ideas: lawsuits, regulation (copyright, antitrust, AI liability), micropayments/revenue‑sharing for training and AI answers, technical blocking of crawlers, or deliberate “poisoning” of AI.
- Others advocate political organizing, switching search engines, boycotting frontier AI, or building/using open local models.
- Strong disagreement remains on whether this shift is net progress, catastrophe, or undecidable; long‑term outcomes are widely acknowledged as unclear.