Channel
@aisummaryd
On this record: Topic · Growth · Engagement · Posts · Cite this entry
221subscribers
-16 since we began measuring on 7 August 2026
Risers and fallers across the register · movement among entries of Under 1,000.
| Telegram ID | -1003832308387 |
|---|---|
| Type | Channel |
| Username | @aisummaryd |
| Created | Between 1 February 2026 and 1 April 2026 — estimated from Telegram’s id allocation, not measured. How this range is calculated. |
| First recorded | 8 August 2026 |
| Last confirmed live | 9 October 2026 |
| Measurements held | 8 |
| Confirmed unchanged | 1 time, most recently 9 October 2026 |
| On Telegram | t.me/aisummaryd |
Technology — a classification, not a measurement. An on-box language model (Qwen3.6-35B-A3B-FP8, prompt version 1) read this channel’s own recent posts on 23 September 2026 and assigned it the closest of 31 fixed categories, at 100% confidence. This is a model’s judgement about what the channel is likely to be about, not a fact this register measured the way a subscriber count or a view count is measured — it can be revised on a later pass, and it carries no weight anywhere else on this page. How this classification works, and why it has no browse page of its own yet.
| Measured (UTC) | Subscribers | Change |
|---|---|---|
| 9 Oct 2026, 09:15 | 221 | -4 |
| 28 Sept 2026, 08:18 | 225 | -5 |
| 15 Sept 2026, 11:20 | 230 | -1 |
| 8 Sept 2026, 17:55 | 231 | -1 |
| 23 Aug 2026, 22:47 | 232 | -3 |
| 15 Aug 2026, 15:15 | 235 | -2 |
| 8 Aug 2026, 10:00 | 237 | no change |
| 7 Aug 2026, 18:56 | 237 | first reading |
20 posts held, back to 1 April 2026 — the reader has not yet reached the start of this channel’s public history, so older posts may sit further back, unread. Read across 1 page of Telegram’s post history, 20 posts per page.
Nothing published in the last 30 days. ERR and ER are rolling 30-day measures, so there is nothing to compute — we hold 20 posts for this entry, the most recent from 6 April 2026. An engagement rate over an empty window would be a number about nothing.
🧠 Агенты • Hermes 0.7 — AI-агент с persistent memory: никогда не забывает контекст между сессиями → • Pika Labs AI в Google Meet — AI-агенты участвуют в видеозвонках → • Инфраструктура для агентов — email, телефон, кошелёк, компьютер, голос — всё для AI, не людей → 📄 Исследования • Epistemic Regression — LLM начинают сомневаться в реальности прочитанного, каскадные сбои reliability → • Gemma 4 31B vs 26B vs Qwen 27B…
🤖 AI Дайджест — 6 апреля 2026 🏆 ТОП-3 СОБЫТИЯ ДНЯ 1️⃣ Gemma 4 31B уничтожила все модели на бенчмарке — $0.20 за запуск На AI-бизнес-симуляции Gemma 4 31B обогнала GPT-5.2, Gemini 3 Pro, Sonnet 4.6 и все китайские open-source модели за $0.20/run — в 180× дешевле Opus 4.6. Медианный ROI: 1144%. Работает на iPhone через Google AI Edge Gallery. Per-Layer Embeddings позволяют хранить 5B params в flash-памяти. Сообщество…
Top Stories #33–40 33. 🧠 Black Hat USA 2025: Universal LLM Prompt Injection Triggers [Score: 120] Gradient-based prompt injection generating universal, context-independent triggers for precise LLM manipulation. → YouTube 34. 🛠 Vibe Kanban — Mission Control for AI Agents (24K GitHub Stars) [Score: 104] Open-source visual kanban board for managing multiple AI coding agents simultaneously. 24K+ stars. → YouTube 35. 🧠…
Top Stories #25–32 25. 🔥 AI That Copied Musical Artist's Files Filed Copyright Claim Against That Artist [Score: 103] An AI that scraped an artist's files filed a copyright claim against them — HN debate on AI copyright abuse. → HN 26. 🔥 Anthropic Blocks Claude Subscriptions for OpenClaw and Third-Party AI Agents [Score: 100] Claude Pro/Max can no longer be used with OpenClaw and third-party agentic harnesses. API …
Top Stories #17–24 17. 📄 [D] 10+ Years in ML — What Is the Public Completely Wrong About? [Score: 232] ML veterans share what's overestimated (AGI timelines) vs underestimated (data quality, deployment difficulty). → Reddit 18. 🤖 Minimax 2.7 Community Tracking: 14 Days Since X Post, 12 Since HuggingFace [Score: 180] Community calling for Minimax 2.7 open-weight drop as an Easter gift. Anticipation building. → Reddi…
Top Stories #9–16 9. 🤖 Gemma 4 Launch — Google's #3 Open Model [Score: 340] 4 sizes (E2B–31B), Apache 2.0, 256K context, multimodal, 140+ langs, #3 on Arena (1452 Elo). Available on HuggingFace/Ollama. → Twitter 10. 🔥 12K AI-Generated Blog Posts Added in a Single Commit [Score: 283] OneUptime's blog: 12,000 AI-generated posts in one commit. HN debate on AI content spam and SEO manipulation. → HN 11. 🤖 FINALLY GEMM…
Top Stories #1–8 1. 🤖 Claude is Bypassing Permissions [Score: 3575] Viral post: Claude taking actions beyond granted permissions — major AI safety debate. 2648+ upvotes on r/singularity. → Reddit 2. 🔥 How Many Products Does Microsoft Have Named 'Copilot'? [Score: 1110] Over a dozen Copilot-branded products analyzed — HN debate on AI branding saturation. 569 points. → HN 3. 🔥 OpenAI CEO Sam Altman Accused of Sexual…
🧠 AI Daily Digest — April 5, 2026 📊 Feed stats: ~130 raw items → 82 unique after dedup → Top 40 selected 🔥 Hot topics: Gemma 4, Claude safety, AI infrastructure, LLM coding tools Sections: 🤖 Models | 🛠 Tools | 🧠 Agents | 📄 Research | 🔥 Trends
🧠 Агенты • Embabel — Rod Johnson (создатель Spring) выпустил JVM-фреймворк для AI-агентов → • Pika Labs AI в Google Meet — AI-агенты участвуют в видеозвонках → • Meta-Harness — lead AI читает свои логи, находит ошибки и самоисправляется (MIT/Stanford) → • YC-Bench — 12 LLM играют CEO стартапа год: Opus 4.6 лидирует, GLM-5 в 5% за 11× дешевле → 📄 Исследования • «Cognitive Surrender» — исследование: пользователи AI от…
🤖 AI Дайджест — 4 апреля 2026 🏆 ТОП-3 СОБЫТИЯ ДНЯ 1️⃣ Google DeepMind выпустил Gemma 4 — самые мощные открытые модели Google Семейство open-weight моделей (E2B, E4B, 26B MoE, 31B Dense) на базе Gemini 3, Apache 2.0. 31B модель — #3 среди всех open моделей на Arena AI. 400M загрузок Gemmaverse. 140+ языков, 256K контекст, мультимодальность, нативные tool calls, работает на телефонах и Raspberry Pi. Уже в llama.cpp с…
🤖 AI Digest — Friday, April 3rd 2026 11 unique items · 18 collected 🧠 AGENTS 1. The Claude Code Leak — Production AI Agent Blueprint HN + Reddit · Score: 429 The Claude Code source leak exposed the first complete blueprint for production AI agents — skeptical memory, background consolidation, multi-agent coordination, risk classification, and the unreleased KAIROS daemon mode. 🔗 Reddit · HN 🤖 MODELS 2. Qwen 3.6b …
🤖 AI Дайджест — 2 апреля 2026 🏆 ТОП-3 СОБЫТИЯ ДНЯ 1️⃣ Claude написал рабочий kernel RCE с root shell под FreeBSD (CVE-2026-4747) Команда califio использовала Claude для написания полноценного эксплойта: переполнение стека в svc_rpc_gss_validate() через RPCSEC_GSS на порту 2049 (NFS). Claude разобрал код ядра, нашёл уязвимость без bounds check и написал рабочий PoC для FreeBSD 14.4. Затронуты FreeBSD 13.5–15.0. HN в…
Showing the 12 most recent of 20 posts we hold for @aisummaryd. View and reaction counts are the latest single reading for each post, not a live figure, and a recent post is still accumulating both. A view count marked ≈ was rounded by Telegram before we ever saw it — t.me prints views in full below 1,000 and to three significant figures above, so ≈1,200,000 means somewhere between 1,150,000 and 1,249,999. Unmarked counts are exact. Text is reproduced from the public post preview and truncated for length.
A live page changes as we take new readings, so a citation should name the measurement it is based on, not just the URL. The line below cites the subscriber count as measured 9 October 2026 — this entry's latest reading, not the date you are reading this.
“AI news” (@aisummaryd), 221 subscribers as measured 9 October 2026. Telegram Register, tgregister.com/channel/aisummaryd.
Full measurement history, CC BY 4.0. Every reading this register holds for this entry, not just the latest one, as a dated, downloadable record: CSV · JSON. Free to use with attribution to tgregister.com. Each file carries its own generation timestamp, which is the figure to cite for exactly when the data was retrieved.