Channel
黑洞资源笔记
群组 @unnamechat
Updated: 2026-10-06
Cashier signals
Ad niches in this channel
Top advertisers in this channel
- 10@zhang_peng_trading@zhang_peng_trading
- 6@pitchscopecn@pitchscopecn
- 4@gamewinbot_bot@gamewinbot_bot
- 3@xuexibaoyang@xuexibaoyang
- 3@gameflux47bot@gameflux47bot
- 3@zehraaarobot@zehraaarobot
- 3@casinov49bot@casinov49bot
- 3@gamenova67bot@gamenova67bot
Activity
Recent posts
- 2026-10-06
Media post
343 viewsOpen in Telegram - 2026-10-04
1809年拿破仑写给马尔蒙元帅的密信在尘封217年后,被GPT-6 Astra仅用6小时成功破译,不仅揭示了战前兵力部署,还补全了中断160年的历史文献空白。 这次破译的技术含量其实不高。密信已知部分明文,又有33个字母对应关系,剩下的就是符号匹配和语料统计——这是计算机最擅长的机械劳动。真正耗时的是人工扫描、分类、校对,AI只是把重复工作做快了。 破译本身…
- 2026-10-04
DeepSeek的“深度思考”模式因频繁出现“啊对对对”等自我辩论的脑内弹幕引发评论区玩梗狂欢,这种前台严谨后台碎嘴的反差萌背后,其实隐藏着AI推理过程外显的技术逻辑。 DeepSeek把推理过程直接摊开给你看,那些“啊对对对”不是情绪爆发,是模型在搜索解空间时的路径标记。人脑思考时也在不断试错和自我纠偏,只是这个过程通常不可见。现在模型把中间步骤暴露出来,…
1,896 viewsOpen in Telegram - 2026-10-04
10月3日,优酷因“全面反对AI、手搓工艺不能丢”的内部会议内容登上热搜,在AI内容生产日益渗透的行业背景下,此举被视为影视平台对纯手工创作的明确站队。 AI和手工的真正边界不在技术,在责任。优酷反AI本质是怕烂剧泛滥后无人买单。AI生成快,但谁对内容负责?编剧署名、演员表演、导演调度,每个环节都有人的判断和取舍,出了问题能找到人。一旦AI大规模介入,责任就…
- 2026-10-04
道士称努力是最没有用的东西 道士静达在访谈中直言“努力是最没有用的东西”,其核心在于区分努力与用心,认为机械重复的肢体劳累远不如身心合一的感受重要。 静达说的不是反对努力本身,而是反对一种麻木状态——身体在动,心不在场。很多人把疲惫当成勤奋的证据,把重复当成进步的信号,这恰恰是最大的陷阱。真正起作用的是注意力的质量,不是时间的长度。 录视频时背稿和感受镜头,…
- 2026-10-04
近日一条“别换成中国老板”的发帖引发广泛共鸣,话题直指国产化管理层替换后福利普遍缩水的群体记忆,将外企与本土企业的管理文化差异推上风口浪尖。 这场争论表面看是福利缩水,实质暴露了两种管理逻辑的碰撞。外企管理者多拿固定薪酬,守住规则就能交差;本土职业经理人常与短期业绩深度绑定,削福利、压工时成了最快的数字优化手段。 问题不在国籍,在于考核周期和激励结构。当企业…
- 2026-10-04
Media post
955 viewsOpen in Telegram - 2026-10-04
【欧洲主权AI再出牌:Kolibri开放权重】 德国AI企业Aleph Alpha正式开源了英德双语MoE模型Kolibri。该模型总参数78B、激活仅3B,支持最高1M上下文,采用Apache 2.0协议。 Kolibri主打欧洲数据主权与严苛合规,研发全流程自主可控。在保持极低私有化部署成本的同时,它采用自研UniBPE分词器大幅优化德语处理效率,并引入…
Research from the blog
All →- 2026-09-30
Crypto Ads on Telegram in 2026: Who Advertises, How Long They Last and What the Rules Say
What our archive of Telegram sponsored messages shows about crypto advertising: how big the niche is, that almost all top advertisers are Telegram accounts and bots rather than websites, how short their lifespans are, what Telegram's ad policy says about crypto, and how to research a competitor yourself.
- 2026-07-10
Backpack vs Phantom on Telegram Ads: Exchange-Wallet vs Self-Custody Leader in 2026
A head-to-head of how Backpack and Phantom advertise on Telegram — the Solana-rooted exchange-and-wallet versus the leading self-custody wallet, their creative and targeting differences, and what the Telegram Ads Spy archive reveals.
- 2026-07-10
Belgium on Telegram Ads: Europe's Strictest Gambling Regime and the Trilingual Split (2026)
How advertisers reach Belgian audiences on Telegram — one of Europe's toughest gambling and ad-restriction regimes pushing operators offshore, a Dutch/French/German language split, and EU-regulated crypto and trading.
- 2026-07-10
Ireland on Telegram Ads: The English-Language EU Bridge and Fintech-Hub Effect (2026)
How advertisers reach Irish audiences on Telegram — an English-speaking EU member inside the UK-Ireland creative pool, a fintech and tech-HQ hub, EU-regulated crypto and forex, and offshore gambling under a tightening licensing regime.
Sponsored impressions seen in this channel · EUR cabinet
Flags — the regions we asked our observer pool for while recording this ad in this channel. That is a label on our request, not a confirmed place of viewing: when no account is available in the requested country the pool silently substitutes another and does not say so. It is not the advertiser's targeting country either. We stopped requesting the region on {date}, so newer creatives carry no flags at all. The channel's main country (by its language) is highlighted in blue; regions under 5% are collapsed into +N.
- ×222026-09-29 → 2026-10-04
- ×122026-09-11 → 2026-10-01
- ×112026-09-16 → 2026-10-04
- ×82026-09-08 → 2026-09-28
- ×72026-09-29 → 2026-10-04
- ×62026-10-03 → 2026-10-04
- ×62026-09-08 → 2026-10-04
- ×52026-09-30 → 2026-10-03
- ×42026-09-16 → 2026-10-03
- ×32026-10-06
- ×32026-10-06
- ×32026-10-04
- ×32026-10-04
- ×32026-10-04
- ×32026-10-04
- ×32026-10-04
- ×32026-10-03
- ×32026-10-03
- ×32026-10-01
- ×32026-09-30 → 2026-10-01
- ×32026-09-25 → 2026-09-26
- ×22026-09-29 → 2026-10-01
- ×12026-10-05
- ×12026-10-05
- ×12026-10-01
- ×12026-09-30
- ×12026-09-29
- ×12026-09-29
- ×12026-09-27
- ×12026-09-27

















