Killed By Claude Report

← Home

Startup

Killed by Claude

What it actually is

Killed by Claude is not an AI product replacing work itself.

It is a small media + database project that curates examples of companies and SaaS categories allegedly weakened by LLMs, then packages that into:

  • short case entries
  • editorial analysis posts
  • a themed tools directory
  • an open-source contribution workflow

The site’s job is basically narrative aggregation: spotting categories that look commoditized by AI, attaching public signals like stock crashes or traffic decline, and turning that into a searchable graveyard for founders, operators, and investors.

So the core product is curation, framing, and audience capture around AI disruption, not frontier model capability.

https://killedbyclaude.com
34Sweating

Current verdict

Anthropic is adjacent, not a direct substitute.

Assessment

Why the score is not higher

Anthropic absolutely strengthens the theme this site feeds on: every stronger Claude release creates more material for disruption commentary.

But that is not the same as replacing the product.

Killed by Claude lives in the layer above the model:

  • editorial taste
  • case selection
  • framing
  • community submissions
  • brand as a public scoreboard of AI casualties

Claude can help generate writeups, summarize evidence, and maybe even draft case pages. But Anthropic is not shipping a public disruption-tracker database or editorial brand aimed at this exact audience.

Why there is still exposure

The weak spot is that this is a fairly light product.

If Claude gets better at web research, long-context synthesis, and computer use, a motivated user could increasingly recreate the site's core output themselves: "find me SaaS categories being hollowed out by AI, give evidence, rank severity."

That makes the workflow easier to commoditize even if the brand itself is not.

Biggest historical hit

Biggest historical hit

Introducing Sonnet 4.6 is the clearest concrete hit.

Its claim of more capable computer use across multi-step web tasks, plus stronger reasoning over messy workflows, pushes Claude closer to doing the exact research grunt work this site depends on:

  • browsing candidate companies
  • checking public indicators
  • synthesizing signals across pages
  • drafting a disruption thesis

That does not kill the site outright.

But it erodes the value of a manually curated tracker when users can increasingly ask Claude to build a custom one on demand.

What still protects them

What still protects them

There is some real insulation, even if it is not a fortress.

  • Brand and tone: "Killed by Claude" is a memorable framing device. People may visit for the voice as much as the facts.
  • Curated canon: a handpicked archive of examples can be more useful than raw model output.
  • Community contribution loop: open submissions and GitHub PRs create a living dataset Anthropic does not maintain.
  • Editorial point of view: the site is selling interpretation and cultural packaging, not just information retrieval.

The problem: these are media moats, not hard product moats.

Useful, yes. Durable against a foundation model giant? Less impressive.

Signals

Web research and synthesis over public company signalsLong-context analysis of many sourcesComputer-use capabilities for browsing and data gatheringNarrative generation from fragmented evidenceGeneral-purpose disruption analysis can be prompted directly

Why this is in the blast radius

Introducing Sonnet 4.6

Anthropic news · 2026-04-11

Inside blast radius
Why it matters

Sonnet 4.6 is described as reaching more useful computer use performance across spreadsheets, web forms, and multi-tab browsing.

That directly overlaps with the startup's lightweight research workflow:

  • inspect public websites
  • gather usage/traffic/company signals
  • compare examples across tabs
  • synthesize a writeup

Killed by Claude is still adding editorial judgment on top, but the mechanical part of the work moves closer to Claude.

Claude Opus 4.6

Anthropic news · 2026-04-11

Inside blast radius
Why it matters

Opus 4.6 is framed around stronger reasoning, long-context handling, and expert-grade research workflows.

That maps well to a product built on assembling fragmented evidence into a coherent disruption case.

If a user can dump articles, traffic anecdotes, earnings comments, and screenshots into Claude and get a plausible disruption memo back, the startup's analysis layer gets partially commoditized.

Introducing The Anthropic Institute

Anthropic news · 2026-04-11

Inside blast radius
Why it matters

The Institute explicitly says it will engage with workers and industries facing AI displacement and report candidly on what Anthropic is learning about the technology's economic effects.

That is not the same product, but it is uncomfortably close at the thematic level.

Killed by Claude tracks AI-driven displacement as public narrative. Anthropic is now institutionalizing analysis of that same topic with more authority and more access.

Partnering with Mozilla to improve Firefox’s security

Anthropic news · 2026-04-11

Outside blast radius
Why it does **not** matter much

This is a demonstration of Claude's capability in software security and vulnerability discovery.

It signals stronger technical reasoning overall, which is directionally relevant. But it does not directly compete with a disruption-tracking editorial site.

So this is background proof of model quality, not a clean substitute threat.

Back to home