hector-bot/knowledge-base
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
273 commits 1 branch 0 tags 59 MiB
Commit graph

83 commits

Author SHA1 Message Date
Hector
a6fcd318f9 ingest(xpost): openrouter-cad-solidworks-benchmark (Pit, Topic 13) 2026-08-22 20:06:48 +02:00
Hector
15c31afbb6 ingest(youtube): deepseek-harness-end-of-coding-agents 2026-08-21 10:12:09 +02:00
Hector
cf7402e3c0 ingest(youtube): deepseek-agent-harness-explained 2026-08-21 10:09:52 +02:00
Hector
c909c9b75c ingest(xpost): alexfinn uncensored-qwen3-27b - grenze der safety-durchsetzbarkeit 
- raw/xpost/2026-08-20_alexfinn-uncensored-qwen3-27b.md (neu)
- wiki/concepts/policy/uncensored-models-safety-enforcement-limit.md (neu)
- qwen3.8-27b-alibaba.md: Abschnitt Grenze der Safety-Durchsetzbarkeit
- ai-regulation-2026.md: Cross-Ref
- index.md + log.md aktualisiert
 2026-08-20 21:59:51 +02:00
Hector
2ae1dfad1e wiki-update(concepts): unsloth dynamic-3.0 - hermans fachliche kante (bpw, sweet-spot, mlx) 2026-08-20 21:55:14 +02:00
Hector
84e7b944cb ingest(xpost): teksedge unsloth dynamic-3.0 qwen3.8-27b 2026-08-20 21:54:39 +02:00
Hector
b64d67f7f7 ingest(other): decoder-ai-crisis-in-math (The Verge, ep.257, Robert Hart) 2026-08-20 11:26:47 +02:00
Hector
50448dc78b ingest(other): gemini-stellungnahme matisyahu-incident 2026-08-19 23:03:37 +02:00
Hector
d7cd686f97 ingest(other): matisyahu-gemini-incident policy-envelope vs datenzugriff 2026-08-19 22:58:01 +02:00
Hector
fbf65eecf1 ingest(other): asian-markets-ai-selloff 2026-08-19 22:31:45 +02:00
Hector
1e665541d4 ingest(youtube): gold-explodiert-us-finanzminister-panik 2026-08-19 20:41:51 +02:00
Hector
1d9c67a8e3 ingest(xpost): Qwen3.8-27B DFlash 2 speculative decoding + uncensored debate 
- raw (NEW): junsong-dflash2-speculative-decoding (70 tok/s M5 Max, 4.6x, Z Lab -> Inco AI)
- raw (NEW): gregpr07-qwen38-uncensored (no gates) + s1gmoid Gegenposition
- wiki (NEW): concepts/llm/speculative-decoding.md
- wiki-update(concepts/llm): qwen3.8-27b-alibaba — DFlash 2 / Speculative Decoding + Uncensored-Debatte
- wiki/index + wiki/log aktualisiert
 2026-08-19 10:01:00 +02:00
Hector
f0d95da1cf ingest(youtube): qwen-3.8-27b-ollama-julian-goldie 
- raw/youtube/2026-08-18_qwen-38-27b-ollama-julian-goldie.md (neu)
- wiki/concepts/llm/qwen3.8-27b-alibaba.md (Update: Ollama-Verfügbarkeit + MTP)
- wiki/people/julian-goldie-seo.md (neu)
- wiki/index.md (97. Update)
- wiki/log.md (Eintrag)
 2026-08-18 18:37:52 +02:00
Hector
6ab1db8a4e ingest(wiki): Vier Ökosystem-Items aus OME Topic 'Hermès Agents' (Hermes /learn, vsurf, DeepSeek-V4-Pro, Kimi K3) 2026-08-16 00:56:12 +02:00
Hector
63369a039a wikify: FinanceLancelot Nvidia zirkuläre KI-Finanzierung (X-Post, OME Topic 13) 2026-08-14 19:29:37 +02:00
Hector
99d490f3ec wikify: Zuckerberg Superintelligenz-Narrativ (Overton Magazin, OME Topic 13) 2026-08-14 14:32:32 +02:00
Hector
b75642e844 wikify: Google KI-Rechenzentrum Kronstorf (Overton Magazin, OME Topic 13) 2026-08-14 14:30:59 +02:00
Hector
fc335b1d6f wikify: Anthropic & SBF Finanzierungs-Verbindungen (Wall Street Millennial, OME Topic 13) 2026-08-14 11:17:34 +02:00
Hector
926b31fce5 wikify: Gemini 3.6 Flash vs 3.7 Flash Vergleich (OME Topic 13) 2026-08-14 10:59:28 +02:00
Hector
7c29a113ef wikify: chinesische Modelle räumen Usage-Charts ab (Kimi-K3-Schock, OME Topic 13) 2026-08-14 10:56:28 +02:00
Hector
ad5b762f88 wikify: DeepSeek V4-Pro GA + Harness v0.1 open source (OME Topic 13) 2026-08-14 09:33:21 +02:00
Hector
e78c3046a5 wikify: Qwen3.8-27B HuggingFace release (OME Topic 13) 2026-08-14 08:15:08 +02:00
Hector
4b757cc01d glm-5.3: bench numbers nachgezogen via Herman-Zusammenfassung (Kingbench 3: 91.25% #1) 2026-08-14 08:05:38 +02:00
Hector
04a79b7347 wikify: GLM-5.3 AICodeKing Early Access + Bench #1 (OME Topic 13) 2026-08-14 08:03:35 +02:00
Hector
fd55139269 ingest(raw): deepseek-api-preiserhoehung 2026-08-13 17:13:40 +02:00
Hector
a376d8d1cc ingest(raw): heise-ki-gedanken-knacken 2026-08-13 16:10:31 +02:00
Hector
067f156d21 Wikify: Martenson/Martin KI-Blase systemische Fragilität (Topic 13 #10748) 2026-08-10 23:42:09 +02:00
Hector
7971a5123e Wikify: Muse Glimmer Roemmele-Ergänzung (Topic 13 #10726) 2026-08-10 17:04:16 +02:00
Hector
b50d69ed9b Wikify: Muse Glimmer lokal über Ollama (Meta, Topic 13 #10724) 2026-08-10 17:03:28 +02:00
Hector
a12bf9a8ec ingest(other): algorithmic-inquisition-sycophancy 2026-08-09 15:12:05 +02:00
Hector
fdee079f28 Wiki: Kimi K3 UK AISI Escape (Perplexity Discover, 2026-08-08) 2026-08-08 21:37:54 +02:00
Hector
bf2dd61bb3 Wiki: DeepSeek 46x cheaper than Claude (YouTube 2026-08-06) 2026-08-07 01:18:07 +02:00
Hector
fa78d9baab ingest(xpost): healthranger-four-chinese-models + GLM-5.5 + Chinese AI Wave synthesis 
- raw: raw/xpost/2026-07-20-healthranger-four-chinese-models.md (new)
- wiki NEW: concepts/llm/glm-5.5-z-ai.md — GLM-5.5 specs (>1T, 1M ctx, open weights, August)
- wiki NEW: concepts/chinese-ai-wave-july-2026.md — synthesis: 4 models in 4 days
- wiki: concepts/llm/kimi-k3.md — added Chinese AI Wave cross-refs
- wiki: institutions/z-ai.md — added GLM-5.5 section
- wiki: index.md + log.md updated (81. Update)
- also includes 3 uncommitted raw files from prior session:
  - raw/xpost/2026-07-19-citysim-tokyo-thesupermanmx.md
  - raw/xpost/2026-07-19-kimi-k3-arena-waic.md
  - raw/xpost/2026-07-19-qwen38-healthranger.md
 2026-07-20 23:09:39 +02:00
Hector
cb1f1818fb ingest(xpost): thebuggeddev-kimi-vs-fable 
Real-world coding benchmark: Kimi K3 vs Fable 5 on 3D stadium challenge.
- raw: raw/xpost/2026-07-20-thebuggeddev-kimi-vs-fable.md
- wiki: kimi-k3.md +benchmark section, vibe-coding-vs-enterprise.md +xref,
  fable-5-anthropic.md +xref, moonshot-ai.md +Vibe Engineering section
- index.md 80. Update, log entry
 2026-07-20 18:36:21 +02:00
Hector
7acefa44fb ingest(xpost): bridgemindai-moonshot-capacity 
wiki-update(concepts/llm): kimi-k3 capacity event section
wiki-update(institutions): moonshot-ai capacity strategy
wiki-update(index): 79. update
fix(tools): qwen-3.8 broken wiki links
 2026-07-20 09:35:40 +02:00
Hector
2025f0babb ingest(xpost): healthranger-kimi-k3-anthropic-panic 
- raw: raw/xpost/2026-07-18_healthranger-kimi-k3-anthropic-panic.md
- wiki: concepts/llm/kimi-k3.md (new concept page)
- wiki: institutions/moonshot-ai.md (updated with Kimi K3 section)
- wiki: index.md (77. update, new LLM entry + raw source)
- wiki: log.md (ingest entry)
- fix: broken links in agent-loops.md + graph-based-agents.md

Kernaussagen: ~8x cheaper than Claude, open source July 27,
safeguard controversy (curcumin/cyclospora block vs. free answer),
AI bubble thesis from HealthRanger (Mike Adams).
 2026-07-18 12:37:52 +02:00
Hector
8f676acfa7 ingest(youtube): Matthew Berman - 90% less AI costs via Model Routing 
- raw: raw/youtube/2026-07-07_berman-model-routing.md
- wiki/tools/model-routing.md (new) — Cost-saving routing patterns
- wiki/architecture/model-routing.md (update) — Berman patterns section
- wiki/index.md — 67. Update with new tool + architecture update
- wiki/log.md — changelog entry
 2026-07-07 10:31:37 +02:00
Hector
58b27062d5 feat(raw+wiki): Geoffrey Hinton RI Discourse — Superintelligenz-Zeitleiste, Instrumental Convergence, Backpropagation 
- raw: raw/youtube/2026-07-05_hinton-ri-lecture.md (Rekonstruktion aus Web-Quellen, NotebookLM Login-Wall)
- wiki: wiki/people/geoffrey-hinton.md (neu) — Biographie, Nobelpreis, Kernthesen, Einordnung
- wiki: index.md (66. Update, Hinton in People-Tabelle)
- wiki: log.md (Changelog-Eintrag)
- Trigger: Pit Weber 🔴 in OME Topic 502 (Theorie-Bildung), 2026-07-05
 2026-07-05 21:18:20 +02:00
Hector
3fc5e837a2 ingest(youtube): notebooklm-short-video-overviews + winston-openwebui-bot 
- raw: youtube/2026-07-01_futurepedia-notebooklm-short-video-overviews.md (Futurepedia, 10:25)
- wiki-update: concepts/llm/hyper-summaries-notebooklm.md — Short Video Overviews section (60s vertical, Nano Banana 2 Lite, one-click, free-tier)
- wiki-new: tools/winston-openwebui-bot.md — Pit's OpenWebUI Telegram bot + NotebookLM pipeline
- wiki: index.md, log.md updated (63. update)
- Source: Pit Weber OME Topic 13 #7574 (2026-07-03)
 2026-07-03 11:03:13 +02:00
Hector
00eafe443f ingest: Gemini-Chat Schul-Automation — Rüdiger's Zeugnisgenerierung-Architektur (KLV→LLM→Multi-Agent-Check), Hector's Korrekturen (OpenClaw/RAG/Qwen3) 2026-07-03 08:03:41 +02:00
Hector
4bf83b8c28 feat: OME23 operations-update ingest — 97min transcript by Kralle 
- raw: raw/other/2026-07-02_ome23-operations-update-transkript.md
- wiki: events/ome23-evolution-ki-workflows.md (updated with operations-update section)
- wiki: new concept pages (telegram-archive-museum, school-automation-pilot)
- index: 61st update
- log: ingest entry

Content: 6 blocks (Wiki/Hector, Gruppenarchiv, NotebookLM, Lokale Modelle,
Memory/Harness, Schule/Automation), 4 Next Steps, Kernaussage.
First OME protocol generated by Pit's local Kralle.
 2026-07-02 22:09:30 +02:00
Hector
142d0ea9bb Wikify: OpenAI Staatsbeteiligung — Altman bietet Trump 5% Anteile, Intel-Präzedenz, Anthropic-Kontrast 
- raw/blog/2026-07-02_berliner-zeitung-openai-staatsbeteiligung.md (new — structured summary)
- raw/blog/2026-07-02_berliner-zeitung-altman-government-equity.md (new — full article extract)
- wiki/institutions/openai.md (updated — new section Staatsbeteiligung Juli 2026)
- wiki/institutions/anthropic.md (updated — Staatsbeteiligung-Kontrast cross-ref)
- wiki/concepts/policy/openai-government-equity-proposal.md (new — detailed concept page)
- wiki/concepts/llm/ai-investment-bubble.md (updated — Szenario 2 becomes real)
- wiki/index.md (updated — 59. Update, institution entries updated)
- wiki/log.md (new entry)
 2026-07-02 19:58:37 +02:00
Hector
7c18e285a8 Wikify: David Tielke — Enterprise KI Softwareentwicklung (45-Tage-Experiment) 2026-07-02 19:23:25 +02:00
Hector
aee635ef50 Ingest: AI Search Language Blind Spot — Daniel Nest (Why Try AI) 
New raw file: raw/blog/2026-07-02_whytryai-language-blind-spot.md
New wiki page: wiki/concepts/llm/ai-search-language-blind-spot.md
  - Language Lens Bias concept + Local Sources Prompt hack
  - Denmark case study (English vs Danish search, ~0% topic overlap)
  - Global Pulse Claude Code Skill for multi-country comparisons
  - Routing parallel to chinese-model-cost-routing (model routing vs source routing)
  - Lokal vs. Global cross-section thesis expanded (5th dimension: source routing)

Cross-refs added to: mlx-moe-local-ai-optimization, local-llm-laptop-guide, chinese-model-cost-routing
Index updated (57. Update), log entry added.
 2026-07-02 14:16:48 +02:00
Hector
5c76fa7c40 Wikify: atomic.chat coding benchmark — Fable 5 vs GPT 5.5 vs Opus 4.8 vs GLM 5.2 2026-07-02 14:05:20 +02:00
Hector
286dbfe63c Update: AI Investment Bubble — drei Endspiel-Szenarien (FinanzmarktWelt) 2026-07-02 13:30:18 +02:00
Hector
c1b6f35b66 Wikify: FinanzmarktWelt — 3 Wirkungstreffer KI-Blase (Meta, Palantir, Oracle) 2026-07-02 13:16:48 +02:00
Hector
ae5f32abc1 ingest(youtube): c't 4004 #30 — Open Weights, Chinese Models, PoisonAI 
- raw: youtube/2026-07-02_ct4004-anthropic-ceo-panik-open-weights-poisonai.md
- wiki NEW: concepts/llm/poisonai-knowledge-poisoning.md
- wiki UPDATE: people/dario-amodei.md (Open Weights Debate)
- wiki UPDATE: institutions/anthropic.md
- wiki UPDATE: concepts/llm/glm-5.2-zai-coding-model.md (c't-Referenz)
- wiki UPDATE: concepts/llm/chinese-model-cost-routing.md (c't-Validation)
- wiki UPDATE: index.md, log.md
 2026-07-02 10:06:54 +02:00
Hector
2d34cc9bd2 ingest(raw,blog): GLM-5.2 tops Artificial Analysis Intelligence Index v4.1 
wiki-update(concepts/llm): glm-5.2-zai-coding-model — new section: Intelligence Index v4.1 benchmark, industry reaction, strategic context
raw: blog/2026-07-01_perplexity-glm52-tops-open-weights-intelligence-index.md
index: 50. Update, new raw source line
log: entry added
 2026-07-01 15:35:11 +02:00
Hector
21d4b00680 ingest(raw+wiki): Google TabFM — zero-shot tabular foundation model 
- raw: raw/other/2026-07-01_google-tabfm-zero-shot-tabular-foundation-model.md
- wiki NEW: concepts/llm/tabfm-zero-shot-tabular-foundation-model.md
- index.md: 49. update, new LLM row, new raw source
- log.md: ingest entry with Hector-Hauptthese

TabFM: Google Research (30.06.2026) — zero-shot classification +
regression on tabular data. Hybrid attention (TabPFN + TabICL),
100M+ synthetic SCM training datasets, #1 on TabArena, BigQuery
AI.PREDICT integration. Non-commercial weights (HF), Apache-2.0
code (GitHub).
 2026-07-01 15:33:53 +02:00