•twitter15877
listening to
timeline
15877
mentions
tracked
tracked
net sentiment
+24%
4998 positive
9734 neutral
1145 negative
volume & sentiment over time
Sep 30 – Oct 14, 2026 · Daily
sources
most discussed
—
active days
15
mentions
Showing 3076–3100 of 4998 mentions
-
• twitter 1w positivegrok bot crews are f**king insane. i gave 6 bots one overnight job and woke up to 612 finished tasks, 47 blocked actions, 89 rejected handoffs and 0 irreversible moves made without me. [the setup was basically this 👇] 1. scout scans the feed and builds the candidate list, but it never gets to decide what actually mo… grok bot crews are f**king insane. i gave 6 bots one overnight job and woke up to 612 finished tasks, 47 blocked actions, 89 rejected handoffs and 0 irreversible moves made without me. [the setup was basically this 👇] 1. scout scans the feed and builds the candidate list, but it never gets to decide what actually moves forward. 2. quant scores what survives and attaches the numbers, but it cannot resize a position or touch the final action. 3. risk reads the evidence and returns PASS, CUT or KILL. if one number shows up without a source, the entire handoff gets bounced. 4. exec is the only seat that can touch the final action, and even that stays locked until a timestamped PASS is sitting there. 5. scribe keeps the source, read time, evidence and result attached to every handoff, so one made-up number can’t quietly travel through the crew. 6. chief keeps all 5 seats moving in the right order, but never does the research, scoring or execution itself. what surprised me is that the crew got more useful once every bot became intentionally incomplete. they can keep working after i close the laptop, argue with each other, reject bad work and finish hundreds of tasks overnight. but when something can’t be undone, the whole system stops. 6 bots can run the night. the last stopline is still me.view source ↗
-
• twitter 1w positiveHow to fuck up your life in the next 30 days (start in October) 1. Deep clean your room until it looks like a sterilized hotel room.🫵 How to fuck up your life in the next 30 days (start in October) 1. Deep clean your room until it looks like a sterilized hotel room.🫵view source ↗
-
• twitter 1w positiveRT @garrytan: Harnesses from labs have an incentive to burn tokens, which means harnesses from startups have a real utility, like how Grep… RT @garrytan: Harnesses from labs have an incentive to burn tokens, which means harnesses from startups have a real utility, like how Grep…view source ↗
-
• twitter 1w positiveHyperFrames just became an app. Describe the video and your agent builds the first cut. then you direct it: circle something on the frame, pin a comment to one word, or just trim it yourself. Works with Claude Code, Codex, or your own agent, and exports in 4K when it's done HyperFrames just became an app. Describe the video and your agent builds the first cut. then you direct it: circle something on the frame, pin a comment to one word, or just trim it yourself. Works with Claude Code, Codex, or your own agent, and exports in 4K when it's doneview source ↗
-
• twitter 1w positiveOpenAI and Anthropic are spending billions making one model bigger @BanburyRoadAI trained 32 models to work as one swarm called Kardashev-0.7, and found that adding more models scales almost the same as adding more parameters i think the next big jump in AI won't be a bigger model. it'll be more of them working toget… OpenAI and Anthropic are spending billions making one model bigger @BanburyRoadAI trained 32 models to work as one swarm called Kardashev-0.7, and found that adding more models scales almost the same as adding more parameters i think the next big jump in AI won't be a bigger model. it'll be more of them working togetherview source ↗
-
• twitter 1w positiveSay hello to Rexie, our mascot for Rex! Superlogical public testing is beginning. We’re working through our mailing list in the order people signed up. Initial testing is limited to macOS but that’ll expand quickly. Welcome, new testers. https://t.co/qNfJnhjnQn Say hello to Rexie, our mascot for Rex! Superlogical public testing is beginning. We’re working through our mailing list in the order people signed up. Initial testing is limited to macOS but that’ll expand quickly. Welcome, new testers. https://t.co/qNfJnhjnQnview source ↗
-
• twitter 1w positive1,500 merged PRs in 5 months. that's what OpenAI got out of Codex once they stopped blaming the model and fixed the harness this is free f*cking gold for anyone still swapping models every time their agent breaks someone mapped OpenAI, Anthropic and LangChain's agent playbooks into one 12-step architecture instead of… 1,500 merged PRs in 5 months. that's what OpenAI got out of Codex once they stopped blaming the model and fixed the harness this is free f*cking gold for anyone still swapping models every time their agent breaks someone mapped OpenAI, Anthropic and LangChain's agent playbooks into one 12-step architecture instead of a pile of buzzwords loop, graph, skill, subagent, harness, one flow trigger → typed nodes → code routes → verifier → human approval, no free-form handoffs every term tied to a decision the system has to make, not just another feature to bolt on agent = model + harness give this to your next agent build before you swap models againview source ↗
-
• twitter 1w positiveLove seeing this! Sheets has always been where people build the little app they need for their own workflow. Gemini makes that much easier now, of course! Love seeing this! Sheets has always been where people build the little app they need for their own workflow. Gemini makes that much easier now, of course!view source ↗
-
• twitter 1w positiveThis guy’s track record: - 3rd day on the team - 3rd demo shipped But in all seriousness, the Exploration Agent is absolutely amazing, and you should try it. This guy’s track record: - 3rd day on the team - 3rd demo shipped But in all seriousness, the Exploration Agent is absolutely amazing, and you should try it.view source ↗
-
• twitter 1w positiveFrom @TheAthletic: The World Gurning Championships is one of many weird and wonderful events held at the Egremont Crab Fair in northern England. Contestants put their heads through a horse collar, distorting their faces before a team of judges. https://t.co/g0YjOBhSN7 From @TheAthletic: The World Gurning Championships is one of many weird and wonderful events held at the Egremont Crab Fair in northern England. Contestants put their heads through a horse collar, distorting their faces before a team of judges. https://t.co/g0YjOBhSN7view source ↗
-
• twitter 1w positiveI genuinely think cursors logo is one of the best logos of all time https://t.co/ZA0yrpep32 I genuinely think cursors logo is one of the best logos of all time https://t.co/ZA0yrpep32view source ↗
-
• twitter 1w positiveDeadline tonight! Take your last opportunity to enter for a chance to see @Lin_Manuel, Paul Rudd, @jimmyfallon and more belt their hearts out for a great cause. https://t.co/4cDoz1n4yy https://t.co/RgVB6LlPYn Deadline tonight! Take your last opportunity to enter for a chance to see @Lin_Manuel, Paul Rudd, @jimmyfallon and more belt their hearts out for a great cause. https://t.co/4cDoz1n4yy https://t.co/RgVB6LlPYnview source ↗
-
• twitter 1w positiveNeed a new Linux Laptop to replace my MacBook Battery lifetime and build quality are my top priorities, what is good in the market nowadays? Need a new Linux Laptop to replace my MacBook Battery lifetime and build quality are my top priorities, what is good in the market nowadays?view source ↗
-
• twitter 1w positiveWarm tones for cool people. Introducing the new Haus Labs Sun Dipped Bronzing Veil. 11 Shades are available now at https://t.co/30VDKdUPj0 and @Sephora. Rylee shares what she's learned through her journey with hormones and wellness, and she hopes to study Traditional Chinese Medicine or holistic health. So yes, what g… Warm tones for cool people. Introducing the new Haus Labs Sun Dipped Bronzing Veil. 11 Shades are available now at https://t.co/30VDKdUPj0 and @Sephora. Rylee shares what she's learned through her journey with hormones and wellness, and she hopes to study Traditional Chinese Medicine or holistic health. So yes, what goes on her skin matters. Rylee wears Sun Dipped Bronzing Veil in Sand.view source ↗
-
• twitter 1w positiveUsing Frontier models like Astra and Opus is like Coaching Kobe. Good motivational speeches about the vision, good spacing, have everyone get out of his way. Using Frontier models like Astra and Opus is like Coaching Kobe. Good motivational speeches about the vision, good spacing, have everyone get out of his way.view source ↗
-
• twitter 1w positiveRT @kanjun: And this is why we built fine-grained permissions into Imbue Studio, so agents can read, label, draft emails without being able… RT @kanjun: And this is why we built fine-grained permissions into Imbue Studio, so agents can read, label, draft emails without being able…view source ↗
-
• twitter 1w positiveRT @kanjun: More and more, I’m replacing my phone with personal apps I built in Imbue Studio. On my phone it looks like a regular home scr… RT @kanjun: More and more, I’m replacing my phone with personal apps I built in Imbue Studio. On my phone it looks like a regular home scr…view source ↗
-
• twitter 1w positiveSolar Mini 4 is free on Nous Portal for the next two weeks. here's how to try it in Hermes Agent: 1. new to Hermes? run hermes setup --portal to sign in to Nous Portal 2. in any Hermes chat, type /model upstage/solar-mini4:free 3. want it as your default? add --global at the end enjoy! Solar Mini 4 is free on Nous Portal for the next two weeks. here's how to try it in Hermes Agent: 1. new to Hermes? run hermes setup --portal to sign in to Nous Portal 2. in any Hermes chat, type /model upstage/solar-mini4:free 3. want it as your default? add --global at the end enjoy!view source ↗
-
• twitter 1w positiveSame weights, 62% under one harness and 33% under Claude Code! Feels like the harness is carrying half the score, and I'm still not sure our benchmarks admit it. Same weights, 62% under one harness and 33% under Claude Code! Feels like the harness is carrying half the score, and I'm still not sure our benchmarks admit it.view source ↗
-
• twitter 1w positiveI’m building my dream football app ⚽️ YOLO26 for ball detection, automatic tracking, and Core ML experiments on IOS. Adding juggling counters, shot analysis, and some cool ball effects. Still a work in progress, but it’s coming together! 🚀 https://t.co/X645VLnAPf I’m building my dream football app ⚽️ YOLO26 for ball detection, automatic tracking, and Core ML experiments on IOS. Adding juggling counters, shot analysis, and some cool ball effects. Still a work in progress, but it’s coming together! 🚀 https://t.co/X645VLnAPfview source ↗
-
• twitter 1w positiveThis makes required checks for sensitive operations part of the type system Especially useful for agent-written code 1. Define required checks 2. Require typed proof 3. Catch missing proof at compile time This makes required checks for sensitive operations part of the type system Especially useful for agent-written code 1. Define required checks 2. Require typed proof 3. Catch missing proof at compile timeview source ↗
-
• twitter 1w positivebtw this pic was from my 19th birthday. would it be helpful for the VCs if I dig up my other birthday photos so you know what to invest in? btw this pic was from my 19th birthday. would it be helpful for the VCs if I dig up my other birthday photos so you know what to invest in?view source ↗
-
• twitter 1w positiveSay goodbye to your 9 to 5 job. You can earn $8k/month if you have: 1. A Laptop 2. A Wifi connection 3. A ChatGPT account (free or paid) Here's exactly how: Say goodbye to your 9 to 5 job. You can earn $8k/month if you have: 1. A Laptop 2. A Wifi connection 3. A ChatGPT account (free or paid) Here's exactly how:view source ↗
-
• twitter 1w positivePay attention to this if you are building with agents. Simulated companies are a big deal for improving agentic products. This is because good agent evals need environments that behave like real companies. Era by Eon can generate a complete company across Salesforce, Zendesk, Slack, Jira, and other systems. Each s… Pay attention to this if you are building with agents. Simulated companies are a big deal for improving agentic products. This is because good agent evals need environments that behave like real companies. Era by Eon can generate a complete company across Salesforce, Zendesk, Slack, Jira, and other systems. Each system behaves like the real vendor, down to rate limits, pagination, and error codes. You can reset the company and rerun the same tasks after changing the model or the prompt.view source ↗
-
• twitter 1w positiveI wanted a charting library that was light and fast, but could also keep up when “a simple line chart” inevitably turns into way more. So that’s what I built! Give it a try! (Or tell your agent to) I wanted a charting library that was light and fast, but could also keep up when “a simple line chart” inevitably turns into way more. So that’s what I built! Give it a try! (Or tell your agent to)view source ↗