entity · entities/Anthropic.md

Anthropic

status: active · created 2026-06-02 · updated 2026-11-08 · 7 sources · #company #ai-lab #foundation-models

AI lab; maker of the Claude model family and the Claude Code harness, and the model base for agents like Manus. Led by CEO Dario Amodei.

In the wiki

Maker of Claude Code (and "Claude Cowork"), cited as a human-agent collaboration harness (The AI Daily Brief — Why Agents Still Need Humans).
Manus is "an AI agent based on Claude" (Making AI Work — Leadership, Lab, and Crowd).
Publishes a benchmarking guide Mollick recommends as a starting point for organizational benchmarks.
One of the labs Every alpha-tests models from.
CEO Dario Amodei is a lead voice in AI and Employment — Replacement vs Expansion (the "half of entry-level white-collar jobs" warning).
Anthropic Research is a subscribed feed for this wiki (auto-ingest, from 2026-11-08). First ingested post: Anthropic — The Components of Agent Memory. It covers how Claude decides what to remember, offline "dreaming," and an evaluation of agent-maintained wikis. It also names products: Claude Managed Agents (persistent memory directories, dreams), Claude Duet, and the 2026 acquisition of Stainless.
Economics team / science usage: Anthropic — How Scientists Use Claude (2026-11-05) analyses 207k research sessions with privacy-preserving classifiers. It extends the Anthropic Economic Index line of work ("economic primitives," Appel et al. 2026). It introduces Claude Science, a specialized interface that connects Claude to fragmented research databases and tools. Over 50% of its sessions are in biomedicine and biology. Its tasks are about 2× longer, more often doctoral-level, and run with high autonomy in over 70% of sessions. The report also names Chat and Cowork as the general-purpose interfaces, which settles the "Claude Cowork" spelling noted on Claude Code.
In-house biology research lab (Life Sciences Team). Anthropic runs its own wet lab and publishes findings "early… to show the scientific community how Claude can be used for discovery." Anthropic — Investigating Vault Particles with Claude (2026-11-04) reports the first 3D structure of a vault particle in prokaryotes, solved in under four months by a team with limited structural-biology experience, using Claude Science for sequence search and an internal multi-agent harness (hundreds of GPUs) for cryo-EM processing. It follows an earlier post on a novel enzyme system (#to-read). Notably, the post documents Claude's errors and states that lab notes were fed back to the ML researchers who train Claude, shaping training data. Contributors: Januka Athukoralage, Peter Yoon, Nicholas Perry, Matthew Durrant; Leonard Rome (vault co-discoverer, 1986) advised as a paid consultant.
Frontier Red Team / inference verification. Anthropic — Progress on Provable Inference for Large Language Models (2026-11-02) releases FlashProve, an open-source (Apache 2.0) zkSNARK prototype that proves a response came from committed weights without revealing them, built largely by Claude and backed by Lean proofs (Provable Inference). Anthropic's Frontier Safety Roadmap had set a public goal of a provable-inference prototype by 2026-09-30; it reports meeting it through a separate project that signs inference traffic and recomputes it on a trusted server. On 2026-10-19 the goal was renamed "inference verification." Co-authors include academic cryptographers (Alessandro Chiesa, EPFL / StarkWare) and Anthropic Fellows. A related Frontier Red Team post, "LLMs are making formal software verification feasible at a much larger scale" (2026-10-22), is a reading lead.
AI for Science program (grantees' guest posts). Anthropic — Teaching AI to Build Scientific Models (2026-11-06) is a guest post on the research blog by grantees Jakob Macke, Daniel Gedon and Stefan Wahl (ELLIS Institute Tübingen / University of Tübingen). It is the first feed item written by outside researchers rather than Anthropic staff, so it reflects the program's grantees' views, not Anthropic's own findings. Reviewed by Xander Balwit, Tom Chivers and Siddharth Mishra-Sharma. See Distributions over Best Fits.

Connections

Distributions over Best Fits — the AI for Science grantees' case for reporting model distributions, not best fits.
Provable Inference — the Frontier Red Team's cryptographic route to verifying which model served a response.
AI Use Tracks AI Capability · AI-Accelerated Science — the Economics team's science findings.
Verification Requires Domain Knowledge — the lesson Anthropic's own lab drew from working outside its expertise.
Claude Code · Manus · Dario Amodei · Codex (the OpenAI counterpart).
Agent Memory · Dreaming (Offline Memory Consolidation) — Anthropic's research framing of agent memory.

Sources

Making AI Work — Leadership, Lab, and Crowd · The AI Daily Brief — Why Agents Still Need Humans · Anthropic — The Components of Agent Memory · Anthropic — How Scientists Use Claude · Anthropic — Investigating Vault Particles with Claude · Anthropic — Progress on Provable Inference for Large Language Models · Anthropic — Teaching AI to Build Scientific Models.

Linked from 10

Anthropic — How Scientists Use Claudeauto-ingested

Anthropic Economics report on 207,049 privacy-preserved research sessions on Claude, Claude Code and Claude Science.

Anthropic — Investigating Vault Particles with Claudeauto-ingested

A result from Anthropic's in-house biology lab: the first 3D structure of a vault particle in prokaryotes, with Claude used at every step — literature and sequence search, DNA construct design, purification-method optimization, and cryo-EM processing.

Anthropic — Progress on Provable Inference for Large Language Modelsauto-ingested

Anthropic's Frontier Red Team releases FlashProve, an open-source prototype that attaches a cryptographic proof (a zkSNARK) to a model's output, showing the response came from a committed set of weights without revealing them.

Anthropic — Teaching AI to Build Scientific Modelsauto-ingested

A guest post on Anthropic's research blog by three AI for Science grantees.

Anthropic — The Components of Agent Memoryauto-ingested

Anthropic Engineering post on how Claude decides what to remember between sessions, and how offline "dreaming" corrects, restructures and prunes that memory.

Claude Code

Anthropic's agentic coding/work harness, cited alongside Codex as a shared workspace for human-agent collaboration.

Codex

Claude Code — the Anthropic counterpart cited alongside it.

Dario Amodei

CEO of Anthropic; cited as a lead voice in the replacement camp of AI and Employment — Replacement vs Expansion.

Manus

An autonomous AI agent "based on Claude" (Anthropic), used by Ethan Mollick as a reference point for what agentic systems can do.

OpenAI

Codex · Anthropic (the rival lab behind Claude Code).

Links to 22

Claude Code
Manus
Dario Amodei
The AI Daily Brief — Why Agents Still Need Humans
Making AI Work: Leadership, Lab, and Crowd
Every
AI and Employment — Replacement vs Expansion
Anthropic — The Components of Agent Memoryauto-ingested
Agent-Maintained Wikiauto-ingested
Anthropic — How Scientists Use Claudeauto-ingested
Anthropic — Investigating Vault Particles with Claudeauto-ingested
Anthropic — Progress on Provable Inference for Large Language Modelsauto-ingested
Provable Inferenceauto-ingested
Anthropic — Teaching AI to Build Scientific Modelsauto-ingested
Distributions over Best Fits
AI Use Tracks AI Capabilityauto-ingested
AI-Accelerated Scienceauto-ingested
Verification Requires Domain Knowledgeauto-ingested
Codex
OpenAI
Agent Memoryauto-ingested
Dreaming (Offline Memory Consolidation)auto-ingested

In the log 5

2026-11-08 auto-ingest

Anthropic — Teaching AI to Build Scientific Models (Anthropic Research guest post by AI for Science grantees Macke/Gedon/Wahl, 2026-11-06; subscribed-feed capture) — unattended; 1 source + 1 concept (Distributions over Best Fits — many models fit equally well; report the range and what failed, because interpretation rests on what was ruled out), both tagged auto-ingested. Kept lean per T5 (SBI, Hodgkin–Huxley and model-space search folded into the source and concept pages; no entity pages for the authors). Updated AI-Accelerated Science (fitting accelerates, understanding doesn't), The Bottleneck Shift (model building moved to AI; interpretation + experiment choice remain), Verification Requires Domain Knowledge (second, informal scope-narrowing route: AI reports its dead ends), Falsification Is Asymmetric (eliminate-the-most-models; answers its sample-frame question), Critical Thinking Is Domain-Bound + Agent Memory (see-also), Anthropic (AI for Science grantees; first outside-authored feed item), Overview, index. Method Reflection: T4 sharpened ("the process of constructing it provides the insight"; candidate: a "Considered and rejected" section in auto-ingest summaries, trialled in this PR); T5 tension (salience vs keep-the-failures, reconciled under "persist what was expensive to work out"); "refuted seeds kept as dead-ends" added to what has held up. No contradictions. Seeds evaluated → neither bears; no callout. 3 needs-your-call (T4 candidate, tension-vs-contradiction judgment, qmd). qmd search still broken (Node ABI) — searched with grep/index.
2026-11-08 auto-ingest

Anthropic — Progress on Provable Inference for Large Language Models (Anthropic Frontier Red Team, 2026-11-02; subscribed-feed capture) — unattended; 1 source + 1 concept (Provable Inference — zkSNARK proof that a response came from committed weights; principle "trust the verifier, not the prover"), both tagged auto-ingested. Kept lean per T5 (FlashProve and zkSNARK folded in as aliases, no entity pages). Extends the software-security domain. Updated Verification Requires Domain Knowledge (scope-narrowing note: a trusted verifier + audited statements shrink the knowledge checking needs; not a contradiction), Control-Flow Integrity (sibling: in-line ~16% vs proved ~70–95×), Allowlist over Blocklist (model approval = allowlist of weights), Falsification Is Asymmetric (Proof-of-Failure certifies effort, not absence), Token Shortage (proving is a ~72–94× compute tax), AI-Accelerated Science (see-also: Claude-built crypto research), Anthropic (Frontier Red Team, roadmap goal, "inference verification" rename), Dario Amodei ("We must pace the frontier"), Overview, index. Method Reflection: T3 counterweight added — shape claims for cheap checking (candidate: verbatim quote per key claim), not adopted. No contradictions. Seeds evaluated → neither bears; no callout added. 2 needs-your-call (scope-vs-contradiction judgment, T3 candidate). qmd search still broken (Node ABI) — searched with grep/index.
2026-11-08 auto-ingest

Anthropic — Investigating Vault Particles with Claude (Anthropic Life Sciences, 2026-11-04; subscribed-feed capture) — unattended; 1 source + 1 concept (Verification Requires Domain Knowledge — you can only catch AI errors where you already have expertise), both tagged auto-ingested. Deliberately lean per T5 salience. The wiki's first worked case of AI-accelerated science: prokaryotic vault structure (phage L05, 96 copies, 9.5 MDa) in <4 months by a team outside its expertise; purification 3 days → 90 min, first model in 34 h; Claude's errors documented (2 weeks lost to a bad protocol; mirror-image cryo-EM map; "we told it to stop improvising"). Updated AI-Accelerated Science (case study + 1 contradiction: physical bench time treated as the cap, but a method substitution cut a physical step ~48× with no lab automation), AI Use Tracks AI Capability (verifiability has two halves — ground truth exists vs operator can read it), The Bottleneck Shift (Claude supplied method ideation + a coenzyme-A hypothesis; also argued against the method that worked), The Human Sandwich (the pattern in a wet lab; the bread must be load-bearing), Agent Autonomy Spectrum (internal multi-agent harness; the one unsupervised improvisation), AI Commoditizes Expertise (borrowed expertise and its bill), Critical Thinking Is Domain-Bound (first field evidence on its open question), Human-in-the-Loop by Exception + Falsification Is Asymmetric (see-also, no source-count bump), Anthropic (in-house biology lab), What's Scarce When Execution Is Cheap (narrows the scarce contribution from judgment toward verification), Overview, index. Method Reflection: T3 sharpened again — review only catches what the reviewer knows, which lands directly on T4's PR-review checkpoint; 4 candidate counterweights recorded. Seed Match Process Weight to Stage and Stakes: third promotion-candidate callout (verdict: analogy only — fourth near-miss; needs a direct source). 3 needs-your-call (1 contradiction, 1 promotion candidate, 1 method question). qmd search still broken (Node ABI) — searched with grep/index.
2026-11-08 auto-ingest

Anthropic — How Scientists Use Claude (Anthropic Research / Economics, 2026-11-05; subscribed-feed capture) — unattended; 1 source, 2 concepts opening the AI & science thread (AI Use Tracks AI Capability, AI-Accelerated Science), all tagged auto-ingested. Updated The Bottleneck Shift (science usage evidence + 1 contradiction: is the what-to-do bottleneck durable? 72% expect AI to match top researchers at ideation by 2030), Individual vs Organizational AI Gains (adoption lags capability; J-curve), Agent Autonomy Spectrum (measured autonomy 23.7% / 64.1% / >70%), Token Shortage (AI cost caps acceleration ~16×), AI and Employment — Replacement vs Expansion (task-level evidence), Human-in-the-Loop by Exception (see-also, not grounding), What's Scarce When Execution Is Cheap, Anthropic (Claude Science, Economics team, Cowork), Claude Code (science usage; Cowork is a separate surface), Dario Amodei ("Machines of Loving Grace"), Overview, index. Method Reflection: sharpened T3 (AI takes analysis + writing, humans keep ideation; that split may not last). Seed Match Process Weight to Stage and Stakes: second promotion-candidate callout (verdict: analogy only). 3 needs-your-call (1 contradiction, 1 promotion candidate, ops). qmd search still broken (Node ABI) — searched with grep.
2026-11-08 auto-ingest

Anthropic — The Components of Agent Memory (Anthropic Research, 2026-11-06; subscribed-feed capture) — unattended; 1 source, 3 concepts opening the agents & memory domain (Agent Memory, Dreaming (Offline Memory Consolidation), Agent-Maintained Wiki), all tagged auto-ingested. Updated Anthropic, Claude Code (MEMORY.md 200-line/25KB cap), Metacognitive Strategies ("what not how" steering), Do It Once, at the Right Place (memory instance), Personal vs Team Agents, Surface vs Deep Structure (wiki beats search on multi-doc Qs), Overview, index. Method Reflection: opened T5 (the wiki as agent memory) with 1 contradiction (CLAUDE.md "not a RAG system" vs wiki+search > wiki-only) + 6 candidate responses. Seed Match Process Weight to Stage and Stakes: promotion-candidate callout (verdict: analogy only, not grounded). 4 needs-your-call (1 contradiction, 1 promotion candidate, T5 responses, ops). Note: qmd search failed (better-sqlite3 Node ABI mismatch) — searched with grep instead.