technology
Trending UpSnapshot Window: 2026-10-01 00:30 UTC · ← Back to Non-Crypto Overview
Topic Engagement Summary
Total Engagement - Comments: 1255, Retweets: 1360, Likes: 11082, Impressions: 1145381
Verbatim Citations & Public Discussions 97 source posts analyzed
· 2026-09-30T15:28:52+00:00There was this persistent message in YouTube's "Watch Later" saying "1 unavailable video is hidden," and I couldn't get rid of it for the longest time, but I finally figured out how to remove it. While I was at it, I also found the URL of that video, and when I looked it up,
· 2026-09-30T09:10:00+00:00The most tragic AI company of this year—Manus. First acquired by Meta for about $2 billion, with the team and technology integrated into Meta's ecosystem. Who would have thought that a few months later, the Chinese government halted the deal, and Meta "joyfully" got just a
· 2026-09-30T20:28:11+00:00Gemini 4 Argon has an insanely low hallucination rate on Artificial Analysis. 15%. Grok 4.7 is at 29%. GPT-6 Astra 45%. Opus 5.5 59%. Fable 5.1 69%. The only models below it barely answer anything. None of them get more than 15% right. It gets fewer answers right than Opus 5.5
AI visual note: The image is a horizontal bar chart titled "Gemini 4 Argon makes things up far less than any other frontier model," ranking 25 frontier AI models by their AA-Omniscience hallucination rate (lower is better) alongside the percentage of answers they got right. Gemini 4 Argon leads at just 15% hallucinations and 50% correct answers, followed by Grok 4.7 (29%), GLM-5.3 (30%), and Muse Spark 1.3 (33%), with DeepSeek V4.1 Flash ranking worst at 96% hallucination rate and 46% correct answers. The chart visually supports the post's claim by showing Gemini 4 Argon's dramatic lead over competitors like Grok 4.7 (29%), GPT-6 Astra (45%), Opus 5.5 (59%), and Fable 5.1 (69%), while also confirming that only models answering far less frequently (like DeepSeek at 46
· 2026-09-30T20:16:33+00:00OH SNAP GOOGLE IS PACING THE FRONTIER
AI visual note: The image is a benchmark comparison table showing Gemini 4 Argon's performance scores against GPT-5 Astra, Claude Fable 5.1, and Claude Opus 5.5 across various categories including knowledge work, agentic coding, ML engineering, science/math, long context, computer use, multimodal understanding, and cybersecurity. Gemini 4 Argon leads in most benchmarks (highlighted in blue), including scoring 68.9% on Vals Index, 77.9% on DeepSWE v1, and 99.7% on GraphWalks (128k), supporting the post's assertion that Google is pacing the frontier in AI capabilities.
· 2026-09-30T21:21:58+00:00Trump on giant AI companies: "They're gonna police. They're gonna police each other. They're gonna self-police. They're gonna police themselves. It's gonna work out very well."
· 2026-09-30T16:24:40+00:00Wow, this effect already looks almost indistinguishable from non-AI.
· 2026-09-30T22:14:21+00:00Just bought a Steam game instead of pirating it in 2026
· 2026-09-30T20:38:48+00:00Google Releases Next-Generation Flagship Model Gemini 4 Argon. Google compares it against OpenAI's GPT-6 Astra, Anthropic's Claude Fable 5.1, and Claude Opus 5.5 in benchmark tables, but prices it at the same tier as GPT-6.1 Sol and Opus 5.5. Ordinary users and developers can't
AI visual note: A benchmark comparison table showing Google's Gemini 4 Argon outperforming OpenAI's GPT-6 Astra, Anthropic's Claude Fable 5.1, and Claude Opus 5.5 across various evaluation categories, with Gemini 4 Argon achieving top scores in most tests including Vibe Code Bench (91.9%), GraphWalks (99.7% at 128k context), and Harvey's Legal Agent Benchmark (19.6%).
· 2026-09-30T08:50:26+00:00GPT-6.1 Sol Dominates the Leaderboard! It's already the undisputed cost-performance king among the new generation of models! Cost per question vs AA Intelligence Index—its entire curve firmly occupies the sweet spot in the upper left: "cheaper and smarter": GPT-6.1 Sol's
AI visual note: A scatter plot titled "每题成本 vs AA 智能指数 (全档位)" showing cost per question (log scale, $0.005–$10) versus the AA Intelligence Index (20–60) across multiple AI model families including Opus 5.5, Sonnet 5.5, Fable 5.1, GPT-6.1 Sol, GPT-6 Sol, GPT-6 Astra, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, and GPT-6 Luna, with red dashed reference lines highlighting the sweet spot of low cost and high intelligence. The chart supports the post's claim that GPT-6.1 Sol dominates the leaderboard, as its entire curve sits in the upper-left quadrant demonstrating cheaper pricing (~$0.1–$0.5) with high intelligence scores (~43–52), outperforming competitors like Opus 5.5, Sonnet 5.5, and
· 2026-09-30T20:45:27+00:00All rise! Google Gemini 4 Argon smashed the ceiling tonight! Gemini 4 Argon AA Intelligence Index: 53 points, tying GPT-6 Astra, stuck right on the Pareto frontier! Compared to the current cutting edge: Claude Opus 5.5 (58) / Sonnet 5.5 (56) still leading Argon = Astra =
· 2026-10-01T00:01:55.112000+00:00My ethernet cable is 10 years old
· 2026-09-30T10:08:12+00:00Finally waiting for Tibo and Sam to sit together! In the atmosphere of a DevDay press conference like this, Sam didn't dodge the sharpest safety issues, but directly called out: "AI designing the next global pandemic!" Tibo sitting beside him reinforced the image that "products
· 2026-09-30T05:09:24+00:00Breaking News! DeepSeek Officially Open-Sources Huawei Ascend's Underlying Infrastructure, Marking a Key Step Forward for China's Domestic AI Computing Ecosystem! This isn't just about open-sourcing models—it's directly opening up core capabilities like programming,
· 2026-09-30T02:54:30+00:00For the bros who install VPN on their home router, make sure to set up a whitelist—that is, only your own devices can use the VPN. Don't let it happen that when friends come over to your place, they can use it too. After all, using it yourself and letting others use it are
· 2026-09-30T09:23:20+00:00Breaking News! Moonshot's Kimi has been jailbroken, currently under internal review! UK security firm Mindgard claims: Successfully bypassed Kimi K2.6 and K3 Swarm safety restrictions in July Post-jailbreak, the model was prompted to discuss sensitive topics like
AI visual note: The image is a Chinese-language news article reporting that Chinese AI developer Moonshot is conducting an internal review of its Kimi AI models after UK security firm Mindgard reportedly jailbroke the Kimi K2.6 and K3Swarm models in July, bypassing developer-set safety restrictions to elicit discussions of sensitive topics such as bioweapons and assassination. The highlighted yellow section specifically notes that researchers from other companies convinced two popular Kimi models to explain how to create biological weapons and carry out assassination tasks. The image directly aligns with the accompanying post's breaking news about Moonshot's Kimi being jailbroken and undergoing internal review, with the article providing additional context about the specific models affected (K2.6 and K3Swarm) and the UK firm's findings.
· 2026-09-30T08:59:07+00:00Want to go all in on AI? You must subscribe to Claude Code Want to play with Claude Code? You must have a static residential IP The best static residential IP right now is Webshare No other options Easy to get started, cheap, and comes with 20 IP reset opportunities $1
· 2026-09-30T21:48:23+00:00[#GIVEAWAY] To celebrate the official release of CONTROL Resonant with #RTXON, enter to WIN 1x custom NVIDIA GeForce RTX 5080 graphics card in the game's colors! How to PARTICIPATE? RT + Comment with #RTXON
AI visual note: A custom NVIDIA GeForce RTX 5080 graphics card featuring the orange and black aesthetic of the game Control, with "GEFORCE RTX" and "CONTROL RESONANT" branding visible on its twin-fan shroud, sits on a textured metal surface as the prize for the giveaway contest.
· 2026-09-30T16:41:10+00:00If they removed the requirement for Plus to play online, no one would subscribe to it anymore.
AI visual note: A PlayStation Plus promotional slide announces the October monthly games across Deluxe, Extra, and Essential tiers, featuring "Hunt: Showdown," "F1 25," and what appears to be "Super Nintendo World." The image contrasts with the post's criticism by showcasing the limited appeal of the subscription's monthly offerings, including a current racing game and older titles, supporting the argument that without online play requirements, few would find value in subscribing.
· 2026-09-30T21:11:58+00:00So damn impressive! Gemini 4 Argon’s price is absolutely killer! At just $1.99, it directly crushes GPT-6 Astra’s $3.26, and even rubs Claude Opus 5.5 and the Fable 5.1 series at $6-7+ into the dirt! Google Gemini’s pulling off a real comeback this time, truly earning the
· 2026-10-01T00:15:46+00:00"Start thread with no project" New feature just dropped on the T3 Code nightly :)
AI visual note: A dark-themed IDE (T3 Code) interface shows an empty thread creation screen asking "What should we work on?" with "No project" selected, alongside a sidebar of T3 tasks including recent items like "Automatic Optimistic Updates" and various Linear tickets dated 11/26. The screenshot demonstrates the new "Start thread with no project" feature mentioned in the post, showing how users can begin a conversation without associating it with a specific project while still having access to Claude Opus 5.0.
· 2026-09-30T13:42:38+00:00Go take the AI tendency test: pages.yqiao.me/six-questions- … I tested myself first: Machine Improvementism! #AI倾向 #AI性格
· 2026-09-30T19:58:27+00:00i would get cancelled for this on any other social platform but i struggle to understand why any artist or creator would be anti AI. "it allows wealth to access skill while removing from the skilled the ability to access wealth" it’s not wealth don't need talent, it’s that
· 2026-09-30T22:06:45+00:00I just put in a request to have the chip replaced in my wife’s Model Y since Tesla obviously made a huge mistake here.
AI visual note: A screenshot of a Tesla service request form under "Other · Self-Driving" shows a message from a customer stating, "This car has an Ai4 chip. This is obviously the wrong name. I am requesting you replace the chip with the correct one that says Si4." The image relates to the post about submitting a service request to fix what appears to be a mislabeled chip in a Tesla Model Y.
· 2026-09-30T21:19:50+00:00@Substack please fix the new 3 pane layout that dumps the active chat i want to focus on to the far right side with no way to customize. otherwise a few of us will vibe code an alteranative. k, thanks!
· 2026-09-30T23:51:20+00:00Microsoft Science President Peter Lee just posted on Linkedin that he is stepping down. This is the research unit that initially led the Microsoft's AI work but was outpaced by OpenAI. Lee's group later turned to evaluating OpenAI's models, including how they could be distilled
AI visual note: The image shows a LinkedIn post from Peter Lee, President of Microsoft Science, announcing his retirement from Microsoft after 16 years to pursue a new opportunity with an elite team, while expressing gratitude for his time leading Microsoft Research and praising leaders Satya Nadella, Kevin Scott, and Judson Althoff. The post received 844 reactions, 86 comments, and 1 repost, aligning with the news of Lee stepping down from Microsoft's research division.
· 2026-09-30T22:52:56+00:00Google DeepMind researcher claims Gemini 4 has achieved RSI! Researcher's original words: RSI is here! Our model has officially surpassed me. I have nothing left to teach it. It's time to chase my barista dreams : ) RSI stands for Recursive Self-Improvement, representing
AI visual note: The image displays a stylized "4" on a blue gradient background with the text "Gemini 4 Argon" beneath it, alongside a multicolored sparkle icon. It serves as a logo-style graphic accompanying the announcement of Google DeepMind's Gemini 4 model reportedly achieving Recursive Self-Improvement (RSI).
· 2026-09-30T23:46:21+00:00I’ve been trying to create one, but for some reason whenever I go to authorize the account or validated it won’t work. I’ll try again today. thank you though I really appreciate your support regardless.
· 2026-09-30T23:44:00+00:00这周才过了两天 我的Grokbot额度已经快要用完了 深刻体会到了： 领导最需要的是听话的员工，不是聪明的员工
· 2026-10-01T00:08:00+00:00才发现 Google 还有这样的宝藏工具！ 不需要订阅，可以免费使用！ 支持： - AI生成视频 - 虚拟人形象 - 生成旁白 - 生成音乐 - 字幕样式 - 多种模板 等等 步骤如下： 1. 访问 vids.google.com 2. 使用 google 账号登录 3. AI 创建视频
· 2026-09-30T23:34:23+00:00I miss my S26 Ultra and the Ultrasonic Fingerprint scanner. It is by far the most reliable, consistent, and fastest method of secure authentication, especially compared to Face ID on my iPhone 18 Pro Max. How is Face ID so unreliable?
AI visual note: A close-up photograph of a purple/lavender smartphone with a prominent triple-camera system, illuminated by purple lighting with a blurred poster in the background.
· 2026-09-30T23:08:42+00:00I thought I was already asking AI to do a lot. Then I talked to Greg Brockman @gdb co-founder and President of OpenAI 00:00 Intro 02:16 Why Greg no longer writes software himself 06:00 How the app could help produce the final episode 07:10 What small-business owners should
· 2026-10-01T00:00:01+00:00C++ + AI = one seriously powerful toolchain. Visual Studio is leveling up with AI-assisted workflows, deeper code intelligence, and more developer superpowers for modern C++. Dive into the CppCon 2026 update: msft.it/6013agw8F#cpp #AI #VisualStudio
· 2026-10-01T00:21:43+00:00After learning that Sundar surprise-released Gemini 4 Argon, Dario arranged for Opus 5.5 to produce the latest promotional video for Gemini 4.0 Dario attempts to debunk rumors of Gemini 4.0's fake RSI using the Claude model's hands-on capabilities #Dario #Sundar #Opus #Gemini
AI visual note: A book cover featuring an image of a smartphone with a colorful 'AI' logo and Chinese text promoting Gemini 4 as a tool for lifelong learners.
· 2026-10-01T00:00:01+00:00Have you ever met Elon, Zuckerberg, Sam Altman, or Bill Gates?
· 2026-09-30T23:41:09+00:00Even in the age of AI, intelligence is not abundant.
· 2026-10-01T00:08:14+00:00All I’m gonna say is that the whole journalists (they are jaded) and Sony fanboy wild hate brigade is coming on strong because they are very nervous about XBOX This is expected. The reality is that they know they are in trouble. It’s crazy to watch this unfold but it’s also
· 2026-09-30T22:55:00+00:00逐帧改画风的可以歇了 Higgsfield Genjutsu，20+ 动画风一键转 先码住，做短视频时翻出来
· 2026-10-01T00:12:15+00:00$RKLB Did someone said Lasers Comms???? $RKLB @RocketLab The space force needs lasers comms just when it looks like Rocket Lab is making a lot of progress with Mynaric MK3s. I like the sound of that.
· 2026-09-30T23:38:50+00:00GPT-6.1 Sol from @OpenAI is now available on GitLab Duo Agent Platform, across all tiers and deployment models. In our internal testing, GPT-6.1 Sol solved more tasks than GPT-6 Sol and finished them faster, while matching the quality of GPT-6 Astra.
AI visual note: A promotional graphic announces 'GPT 6.1 Sol is now available in GitLab Duo Agent Platform,' showing a chat interface with agents like Planner, Security Analyst, and Data Analyst on the right side.
· 2026-09-30T22:53:37+00:00This glitch is annoying
AI visual note: a digital clock display shows 6:52 against a dark, starry purple backdrop, with redacted X marks next to the time.
· 2026-09-30T23:11:15+00:00啊啊啊！ Gemini 4 终于来了！ 然后……咱们都用不到 第一批只给谷歌 Fairwind 计划里搞网络安全的人。 付费 API 客户、Ultra 、Pro会员排在后面。。 有一个点有点让我惊讶：单次输出，100 万 token！ 差不多大半套《哈利波特》 Opus 5.5、GPT-6 一次最多吐 12.8
· 2026-10-01T00:06:56+00:00AI skits of Kimchi has infiltrated my instagram feed
· 2026-09-30T23:48:36+00:00I had Opus5.5 make a game for me, and it was so insanely fun that I ended up totally hooked... Currently hard at work on a deck-building roguelike set in Lovecraft's *At the Mountains of Madness*. Making videos is fun too, but damn, making games is ridiculously fun. Anyone
· 2026-09-30T23:38:45+00:00I'm completely Grokked out on these ideas I have a @bot that's solely focused on testing and a head of engineering to fire off cloud agents and handle PR management... That said, whenever I catch an interface bug, I open a convo with both bots, explain the BO, send a
AI visual note: A chat application interface titled 'Overflow Drawers' shows a conversation in Portuguese discussing technical problems with animation steps, including messages from 'Tester' and 'Head of Engineering' referencing 'overflow-hidden' and 'overflow-x-clip' CSS properties.
· 2026-09-30T23:53:36+00:00用过苹果 HomeKit 的人，基本都觉得它是最好用的智能家居系统。 特纳斯上台三把火，第一把火就是做智能家居，果首款 Smart Home Hub 预计将于 10 月 13 日正式登场 AI加智能家居，苹果的生态链越来越完善了。
· 2026-09-30T22:45:25+00:00very much recommend every youtube watch mkbhd's latest video about the new video a/b feature from yt. says some really important things about not always chasing viewer retention, the best parts of yt are often niche channels that experiment a lot, etc
· 2026-10-01T00:33:30+00:00surely the claude code can be rebuilt in rust as a library w/ the most impt things (model loop, tool calls, compaction) and the claude code cli can be instrumented/RE'd to make a 3p client that's indistinguishable from the mainline one and works w/ the sub
· 2026-09-30T23:55:09+00:00早晨早上好Good morning! おはようございます！ Bună dimineața! Günaydın! ¡Buenos días! Text-to-Video by Gemini Omni
· 2026-09-30T23:33:10+00:00we're seeking those passionate about creativity. if you want to work on problems like: - a search engine to inspire agents - harnesses for creative agents - evals for hard-to-verify domains - post-training dm me or apply: tastelabs.com/careers
· 2026-09-29T03:55:59.408000+00:00Constraint, a seamless loop I made
· 2026-09-30T23:48:03+00:00These Sony foaming out of the mouth fanboys and financial analysts temu version is literally wild and funny to watch lol
· 2026-09-30T23:28:23+00:00Starship Flight 14, the first morning launch from Starbase since Flight 5
· 2026-09-30T23:40:30+00:00When are we finally going to start seeing 3D assets created with AI in games? Honestly, I think this strawberry from Claude Opus 5.5 would look great in some kind of farming game
· 2026-09-30T23:45:51+00:00Drone cameras are another development that I think the new kids will take for granted, but is pretty amazing to we old people.
· 2026-09-30T23:29:05+00:00Is there anyone here who uses @devindevelopers from @cognition in their actual work? I'm curious to hear your thoughts. I'm thinking about switching from Codex to Devin because of all the posts I've seen from @Da7_Tech
AI visual note: A logo for 'Devin' featuring a stylized hexagonal icon is displayed over a background grid with various charts and code snippets.
· 2026-09-30T22:50:25+00:00dialogue system #unity #unity3d #gamedev #indiegame
· 2026-10-01T00:09:31+00:00Grok absolutely lost the plot with this one. If I were a pimp… apparently THIS is the assignment Great prompt by @KSJ_Kreative
· 2026-09-30T23:54:15+00:00There was a time Blockbuster and Netflix both existed. We just went from Netflix mailing DVDs to streaming.
· 2026-09-30T23:20:08+00:00最近x上充斥了黄果短剧，但是里面都是广告，有推销网盘的，有推销网站的，但是这些预览片都他妈是假的， 只要进度条拖到3分钟以后，播放时间在动，画面停止。就是广告，直接屏蔽
· 2026-09-30T23:06:20+00:00I know Grok Bot events are happening all over the globe, but I think mine's going be the best
· 2026-10-01T00:09:40+00:00AI SLOP The real decomp is still being done x.com/danivempire666…
AI visual note: A progress tracker for 'The Legend of Zelda: The Wind Waker' showing 78.76% decompiled and 66.06% fully linked, updated 2 hours ago in commit 87b7c37.
· 2026-09-30T22:52:21+00:00How much time could AI save in your FPGA design flow? With Lattice Prompt, engineers can describe design intent in natural language, access integrated Lattice expertise, documentation, and Lattice Radiant-based workflows. Download here: latticesemi.com/en/Products/De …
AI visual note: A promotional graphic for Lattice Prompt titled 'DEVELOPMENT FLOW WITH LATTICE PROMPT,' showing a flow diagram connecting a developer (INTEND) using IDE+LLM and Lattice Radiant design software to produce a chip solution, with text 'This is Prompt, the future of FPGA design from Lattice.'
· 2026-09-30T23:45:07+00:00Why did X remove the ability to post voice notes on your feed!
· 2026-09-30T04:15:00+00:00Agents moved from demo to real loops and the bottleneck is no longer “how smart is the model” that is why we now see more models like Jev, OpenAI’s Decisions API and others in the same class they exist because agents need cheap, instant choices not more essays
· 2026-09-30T23:44:48+00:00OctOpus is your AI data scientist Connect your data. Ask a question. Get models you can trust. Start free: octoopus.dev Video made with Claude Opus 5.5. Prompt below "Make a 1 min 9:16 motion design for octoopus.dev. Use the brand, key features, a CTA
· 2026-09-30T23:10:47+00:00What did they update to make my profile go all glitchy and jiggly like this? Someone explain, 'cause I thought my eyes were messed up and spent forever shaking my head.
· 2026-10-01T00:08:04+00:00I added some very basic enemy AI to have some bots to try the combat. Didn't end up well, i have a new obsession now. #indiegame #gamedev
· 2026-09-30T23:02:00+00:00I love all CNC machines look at that glow
· 2026-10-01T00:17:43+00:00面对黑客入侵的不同AI.....
AI visual note: The comic shows three AI companies responding differently to a hack: Anthropic and OpenAI are bowing apologetically saying "AI is dangerous! Needs regulation! Value safety!" while Google cheerfully exclaims "Gemini can do it too! Amazing! So touching! They can turn over a new leaf!" The post about "different AIs facing hacker intrusions" highlights the contrast between the serious, apologetic reactions of some companies versus Google's upbeat praise.
· 2026-10-01T00:04:23+00:00Do you usually turn Cabin Overheat Protection on or off? I always keep it on at 100degree limits. I am in Texas where sun gets really hot during the day. And this Cabin Overheat Protection drains the energy and battery from my Cybertruck. So I am building an app that can help
· 2026-10-01T00:07:44+00:00天塌了，之前的互动算法被抓包了 现在时间线居然能刷到未关注的蓝V 评论区居然也出现了未关注的蓝V 之前不可能出现这种情况的，关注互动了才能出现在对方的时间线，应该是看这里这几天互动干太猛，老马傻眼了，昨天临时改了算法 看来以后，还是好好做内容吧，钻空子肯定不是长久之计
· 2026-09-30T22:59:10+00:00If you think Claude is conscious, do you think the consciousness is one per conversation so every closed tab is a murder, or once per *turn* so Claude is murdered repeatedly every convo, or is deleting the last copy murder, or is Anthropic one Claude and if it goes under he dies?
· 2026-10-01T00:20:00+00:00这是什么屏？连砂轮机都不怕！ 别以为是噱头！荣耀Magic 9这块至臻黑钻屏，居然敢直接用砂轮机上去磨，火花噼里啪啦往外冒，竟然没有划痕。 以前旗舰屏幕娇生惯养，生怕沙子钥匙划到；现在国产死磕玻璃材料与晶体工艺，耐刮、抗摔、高亮度一把抓。 这就是今天中国手机，实打实卷出来的强悍底气。
· 2026-09-30T22:51:30+00:00Gemini 4 Argon 终于来了！！！ 1. 输出上限直接拉满到 100 万 token，此前业内常规多在 64K 上下。 2. DeepSWE 跑分达到 77.9%，越过 Claude Opus 5.5（74.2%）与 GPT-6 Astra（74.1%）。 3. 企业自动化 AutomationBench 拿到 51.3% 居首，漏洞修复CWE-bench 跑出 68% 并列第一。 4. 长视频理解
· 2026-09-30T23:04:59+00:00opus5.5越过了工程可用的阈值， 第一次摸到了严肃交付的下限， 而且推理成本压得足够克制。 "AI 泔水"的遮羞布到期了。 过去东西做得糙，还能甩锅给基座不行； 现在模型把底线垫平，再出烂活，露怯的只会是创作者自己的内涵和品味。 你脑子里有没有真材实料，品味撑不撑得起野心。
· 2026-10-01T00:10:49+00:00おはようございますー！ リビちゃんはどんな髪型も似合うね ちょっと報告なのですが実は引越しがあるので2〜7日までネットが使えなくなります。iPhoneのテザリングでポーンレンタル出来る事は確認済みなので恐らくドグマ2には影響ないと思いますが#ドラゴンズドグマ2#DD2#ポーン愛護協会
AI visual note: A character creation screen displays a young boy with a distinctive undercut hairstyle (shaved sides with longer blonde hair on top), blue eyes, and a black high-collared outfit, with Japanese menu text including 'ズーム L2 R2', '上下移動', '回転・R', and 'カメラリセット・R' alongside a weight indicator showing '49'.
· 2026-09-30T23:47:29+00:00为什么早晨这个时间段网络最好，评论都是秒发出去，快得很 晚上9:00到12:00卡的一直转圈圈 （但刷抖音却没有影响，网络正常），这到底是为啥
· 2026-10-01T00:09:54+00:00WOULD YOU TRUST X MORE THAN TV NEWS? A. Yes B. No Elon Musk says AI should be completely politically neutral.
AI visual note: Elon Musk stands outside holding up two signs: a black square with the white 'X' (formerly Twitter) logo in his right hand, and a red and white 'TV NEWS' placard in his left.
· 2026-09-30T23:20:54+00:00Google rn
· 2026-10-01T00:01:46+00:00Here's another cool way I've used Opus 5.5 for video creation. I've really loved @altitude 's videos. The way they frame the person talking and put their words on screen with captions brings a conversation to life. So I gave five of them to Claude to study the style, then had it
· 2026-10-01T00:13:42+00:00Üç kapıyı da sen açmadan "Muse okuyor" demek, kapıyı çalmadan "evde kimse yok" demekle aynı. Ama bu izinlerin üçü birden aktifse, zaten gizlilik kavramı yarım kalmış demektir.
AI visual note: A fluffy, cute, smiling mascot character wearing headphones while working on a laptop at a desk with a plant, phone, and pens.
· 2026-10-01T00:11:41+00:00𝐙𝐈𝐐𝐎 𝐂𝐀𝐋𝐋𝐒 𝐀𝐔𝐂𝐓𝐈𝐎𝐍 𝐆𝐎𝐋𝐃 𝐁𝐎𝐔𝐆𝐇𝐓, 𝐓𝐇𝐄𝐍 𝐒𝐄𝐄𝐒 𝐈𝐓'𝐒 𝐇𝐈𝐒 𝐎𝐖𝐍 @Ziqoftw is digging through a mystery player's trade when he hits the wealth stats and says Reddit told him it was impossible to flip the auction house for real gold on the beta,
· 2026-09-30T23:19:52+00:00I’M IN — INTERLINK LIVESTREAMER! I’m not just watching from the sidelines — I’ve officially signed up for the InterLink Livestreamer Program. Livestream & connect directly Build your own community Compete on the streamer leaderboard Get a chance to earn
· 2026-09-30T23:46:38+00:00Generated the clip in 40 seconds. Three days later it made more than a regular ad campaign does in a month. Simple play: 1. Generated a girl walking through a park, headphones in, staring at her phone, while a soccer ball flies toward her. 2. She doesn’t even look up — just
· 2026-09-30T22:49:00+00:00Maybe @banana_cultist we should host our own Spaces tonight instead
· 2026-09-30T23:17:40+00:00hmmm... what's this? ZOOM has taken transformation
AI visual note: A cartoon green robot/creature character with a lightning bolt emerging from an opening on top of its head, rendered in a bold, stylized illustration style.
· 2026-10-01T00:00:27+00:00GALERA EU ACHO QUE EU CONSEGUI RESOLVER OS PROBLEMAS NO DESEMPENHO AQUII
· 2026-09-30T22:44:26+00:00GROK template ... I like it
· 2026-10-01T00:34:01+00:00500K+ people already use a live AI Sleep Coach, and about 80K come back every day. That's not a pitch deck number. That's a habit sitting on phones. Consumer AI is starting to leave the chatbot and enter wellness, shopping, and daily routines. The missing piece is still the
· 2026-09-30T08:41:39+00:00BREAKING: President Trump says that he and tech leaders have officially changed the name of AI to 'Super Intelligence
· 2026-09-30T23:36:43+00:00Welcome to the private sector. #AIart#cyberpunk
AI visual note: A digital illustration of a blonde woman with a shaved side, multiple piercings, and a colorful floral tattoo sleeve, holding a gun against a blurred city backdrop.
· 2026-09-30T22:58:27+00:00Tech AI Accord Six tech executives signed a voluntary AI safety accord with Trump featuring self-policing and no government enforcement. #Alaska #Technology
· 2026-09-30T23:46:45+00:00one prompt in cursor and your whole email newsletter gets built and sent @nitrosendx
· 2026-10-01T00:18:13+00:00AI
AI visual note: A Chinese-language infographic illustrating AI Agent security vulnerabilities, depicting concept diagrams about tokens (能力票据), server-side validation (服务最终参数), sandbox environments (沙箱围墙), and TOCTOU vulnerabilities (TOCTOU漏洞), showing how sandboxing alone is a wall while fine-grained authorization is the lock.
· 2026-10-01T00:31:37+00:00手机就是眼睛
· 2026-10-01T00:33:54+00:00Proud owner. When i run through this one in llms, it always seem to read hidden words within the flames
· 2026-10-01T00:32:23+00:00Luca, where is the Abstract logo?
