TT Claude Memory

project_susskind_end_of_lawyers_indexed_2026_05_01

Properties1
descriptionRich-purchased Kindle 2026-05-01; targeted-read against ζ-Q15 + ζ-Q16 + InheritKitBench scorecards; 7 syntheses validate locked picks; NO reversals; Acquirer-narrative + Standards-track criteria gain 0-1pp via Susskind-grounding uplift on L4M-decomposed options

3 min read

Indexed: ~/off-github/library/indexed/susskind-end-of-lawyers/ 2026-05-01T15:09 BST.

Source: Richard Susskind, The End of Lawyers? Rethinking the Nature of Legal Services (Oxford University Press, 2008 hardback / 2010 paperback). 281 pages. Rich-purchased Kindle EPUB + PDF.

Reading: TOC+targeted; deep-read Chapter 2 §2.1 (5-step evolution model) + §2.5 (decomposing + multi-sourcing 12-source taxonomy + main-contractor framing) + Chapter 4 §4.6 (legal open-sourcing prediction). 9 verbatim quotes extracted.

Rich expectation: “old book so do not expect any dramatic results”. CONFIRMED ACCURATE. NO scorecard reversals; 7 syntheses validate previously-locked picks.

7 syntheses:

  1. Π.θ ★ L4M 4-stage decomposition validates Susskind §2.5 decomposition principle — exactly what Susskind 2008 advocated, applied to AI-agent architecture rather than human-firm sourcing. Confidence: HIGH.

  2. Σ.η ★ Hybrid main-contractor framing IS direct application of Susskind §2.5 — IAS = main-contractor brand; Word/Clio/LEAP connectors = multi-source suppliers; L4M Stage-3 SMT + per-matter signoff + 10% AI-judge sample = quality systems; Temporal = project management. Confidence: HIGH.

  3. InheritKitBench is Susskind’s “open-sourcing” prediction, exactly fulfilled — 2008 Wikipedia-model prediction for legal materials = LegalBench (Stanford NeurIPS 2023) = TT InheritKitBench v0.1 (planned 2027). Confidence: HIGH.

  4. TT’s product family hits all 5 of Susskind’s evolutionary steps simultaneously: INHERIT v2 standard = standardised; IAS = systematised; InheritKit MCP + 9 Skills = packaged; OpenInherit consumer Phase-1.5+ = commoditised. Defensible spread across 4 of 5 steps (TT explicitly does NOT compete in bespoke). Confidence: HIGH.

  5. Susskind §4.9 “embedded legal knowledge” framing validates IK MCP + 9 Skills cross-vendor architecture — AI vendors embed TT IP within their workflows. Confidence: HIGH.

  6. Susskind doesn’t cover LLM-era specifics (GPT-3 = 2020; Claude = 2023; L4M = Nov 2025; Anthropic Skills = Dec 2025) — but his strategic framework ABOVE the substrate level is substrate-agnostic. We adopt Susskind’s framing without inheriting his pre-LLM tactical assumptions. Confidence: HIGH.

  7. Acquirer-narrative criterion (#14) + Standards-track criterion (#11) gain 0-1pp Susskind-grounding uplift on L4M-decomposed options + main-contractor architectures + open-sourcing-style benchmark designs. Margin Π.γ vs Π.θ remains approximately 0.04 (becomes ~0.05 with uplift; below 0.20 threshold either way; Rich-pick Π.θ ★ unchanged). Margin Σ.η vs Σ.θ same. Confidence: HIGH.

Out-of-scope (read TOC-only): Chapters 1, 3, 4.1-4.5, 4.7-4.9, 5, 6, 7, 8. See exclusion-log.md.

Re-read trigger conditions:

  • Chapter 5 (in-house lawyers): TT pivots to in-house counsel as primary buyer (NOT planned).
  • Chapter 7 (access to justice): OpenInherit Phase-1.5+ public-release planning sprint.

Spike file updates:

  • ~/off-github/library/projects/inherit/T-ai-agent-stack-spike-2026-05-01.md v1.0 → v1.1 (Susskind §2.5 + §2.1 + §4.6 grounding)
  • ~/off-github/library/projects/inherit/T-partner-firm-legacy-migration-spike-2026-05-01.md v1.0 → v1.1 (Susskind §2.5 main-contractor grounding)

No new richard-tasks. No new amendments. No reversal of Π.θ ★ or Σ.η ★ Claude pick.

Graph View

Backlinks