Sure, I may have just sung praises for ChatGPT, an AI in its own right. But I can’t deny that AI is making certain parts of our lives worse, whether by eroding our critical thinking skills, cutting jobs, or hogging precious resources like water, electricity, and hardware (and jacking up prices for everyday folks in the meantime). Then there is the enshittification of the programs that use AI to supposedly make improvements.
Take Google.ai, for example.
Right now, it is a free, ad-free service that delivers answers from Google’s AI. But you can bet that it won’t stay that way for long, and the folks at Boing Boing decided to ask it exactly that. They asked Google.ai to give them its own enshittification timeline, and it delivered:
Timeline of Google.ai Enshittification
Stage 1: User Seduction (Current – 2026)
In this stage, Google prioritizes providing immense, “surplus” value to users at its own expense to achieve massive adoption and lock-in.
- The Goal: Build habit and dependency.
- Example (2025): Google offers AI Overviews (formerly Search Generative Experience) for free, prioritizing clean, fast, and comprehensive summaries that resolve user queries in fewer steps—averaging 2–3 queries per session compared to 5+ in traditional search.
- Sign of Decline: Even during this “golden age,” publishers are already seeing a 50–70% drop in click-through rates as Google shifts toward “zero-click” results.
Stage 2: Business Optimization (2026 – 2028)
Once a critical mass of users is “locked in,” the platform pivots to serve its corporate and advertising partners, often degrading the user experience to do so.
- The Goal: Shift surplus value from users to business customers.
- Example (2026 Prediction): Sponsored AI responses become the default. Instead of the most accurate answer, the top-level AI summary is “grounded” in content provided by paying partners (e.g., specific retailers or service providers).
- The Decay: Conversations with AI become “lead generation” opportunities. As of late 2025, Google has already confirmed that ads are coming to “AI Mode” with no opt-out, turning personal conversational details into behavioral profiling data for advertisers.
Stage 3: Extractive Decline (2029 and Beyond)
In the final stage, the platform aims to maximize short-term profits for its shareholders, impacting both users and business partners.
- The Goal: Capture all remaining surplus value.
- Example (2029 Prediction): High-quality AI results are placed behind a Gemini Premium paywall. The free version becomes filled with low-quality content, misinformation, and low-fidelity ads.
- The Collapse: Ad prices increase while effectiveness decreases due to bot fraud and low-quality content. Experts predict that low-quality content will make up 90% of the internet by 2026.
One good thing about AI is that once it’s fed a certain dataset, it can’t deviate from the conclusions it draws from that. Like the AI vending machine that was talked into communism and gave away $1000 in merchandise. Even Grok has been known to fact-check MAGA at times. So when Google.ai predicts just how its creator will, well, make it shittier — based on the creator’s past and current actions — it’s just amusing, is it not? 😀
