•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 3526–3550 of 4998 mentions
-
• twitter 1w positiveThe past 7.5 months have been crazy. What was supposed to be a 4 month long house project ended up taking twice as long. Been living with the wife and kids at my in-laws while we fix some things at our house. Seeing my kids reactions to our house today was priceless. Got another few weeks of moving and fixing up th… The past 7.5 months have been crazy. What was supposed to be a 4 month long house project ended up taking twice as long. Been living with the wife and kids at my in-laws while we fix some things at our house. Seeing my kids reactions to our house today was priceless. Got another few weeks of moving and fixing up the house but we’re close. Good reminder of why we do what we do. 🙏view source ↗
-
• twitter 1w positiveMost people think an AI agent is just a smarter chatbot, it’s not. An agent perceives, remembers, reasons, plans, acts, observes and loops until the goal is achieved. That’s the real shift from AI that answers to AI that works. I put the anatomy into one simple diagram 👇 10xme — AI that actually works for you. 🔗http… Most people think an AI agent is just a smarter chatbot, it’s not. An agent perceives, remembers, reasons, plans, acts, observes and loops until the goal is achieved. That’s the real shift from AI that answers to AI that works. I put the anatomy into one simple diagram 👇 10xme — AI that actually works for you. 🔗https://t.co/I4zR8aXpsxview source ↗
-
• twitter 1w positiveI am turning your boring features into one that visitors love to engage with. Designed for https://t.co/zGb7TgtjZg by @weswinder Time to animate them, and believe me, you're going to love it. https://t.co/xPr4REl0Zz I am turning your boring features into one that visitors love to engage with. Designed for https://t.co/zGb7TgtjZg by @weswinder Time to animate them, and believe me, you're going to love it. https://t.co/xPr4REl0Zzview source ↗
-
• twitter 1w positiveI like this better than a scatter plot https://t.co/F2S9QJffkD I like this better than a scatter plot https://t.co/F2S9QJffkDview source ↗
-
• twitter 1w positiveI still cannot believe something that dumb is transforming my life to this extent, I spent the day happy and I'm still happy it feels like I'm running out of cool ideas because I just keep throwing tasks at it and it keeps going I still cannot believe something that dumb is transforming my life to this extent, I spent the day happy and I'm still happy it feels like I'm running out of cool ideas because I just keep throwing tasks at it and it keeps goingview source ↗
-
• twitter 1w positiveCancelled my 200$ codex plan, switched to Claude code 100$ plan. I can use opus 5.5 for almost 1 week to hit the token limit, while my 200$ plan can only stand for a couple of days. Also opus 5.5 high is > Astra extra high for my application. Opus 5.5 max is even better than high. I hope gpt catches up soon. Consum… Cancelled my 200$ codex plan, switched to Claude code 100$ plan. I can use opus 5.5 for almost 1 week to hit the token limit, while my 200$ plan can only stand for a couple of days. Also opus 5.5 high is > Astra extra high for my application. Opus 5.5 max is even better than high. I hope gpt catches up soon. Consumer benefit from competition!!!view source ↗
-
• twitter 1w positiveA really good post on creating videos with Opus 5.5 They hit the nail on the head for why frameworks like @HyperFrames_ are important for production ready videos and utilizing tools to improve the overall output instead of just playing around with one shot outputs A really good post on creating videos with Opus 5.5 They hit the nail on the head for why frameworks like @HyperFrames_ are important for production ready videos and utilizing tools to improve the overall output instead of just playing around with one shot outputsview source ↗
-
• twitter 1w positivethe best part of this Jev Engineering loop is that nobody had to tell it when to stop. most autonomous systems are good at starting work. stopping is harder. too early → bad result. too late → wasted compute. here the termination condition is part of the decision architecture. each turn: run agents → score out… the best part of this Jev Engineering loop is that nobody had to tell it when to stop. most autonomous systems are good at starting work. stopping is harder. too early → bad result. too late → wasted compute. here the termination condition is part of the decision architecture. each turn: run agents → score output → keep / reject → update state → ask “goal met yet?” the target was 0.85. turn 18 reached 0.93. Jev returned: stop. 5,400 agent runs finished. 0 questions sent to the human. that’s a much more useful definition of autonomy than “the agent can use tools.” Jev Engineering gives the system a way to know when enough is enough.view source ↗
-
• twitter 1w positiveExpo EAS is dead You can now run free unlimited builds on GitHub runners with React Native Deploy https://t.co/vSNT6o0Lch Expo EAS is dead You can now run free unlimited builds on GitHub runners with React Native Deploy https://t.co/vSNT6o0Lchview source ↗
-
• twitter 1w positiveMy prayer today: Dear God, Thank you for answering my prayers. Well some of them at least. Help me with the rest please. Amen. My prayer today: Dear God, Thank you for answering my prayers. Well some of them at least. Help me with the rest please. Amen.view source ↗
-
• twitter 1w positiveHOLY SHIT WE ARE FUCKING WINNING THIS TIME!! GPT USERS REJOICE!!! 1.2k reposts btw. HOLY SHIT WE ARE FUCKING WINNING THIS TIME!! GPT USERS REJOICE!!! 1.2k reposts btw.view source ↗
-
• twitter 1w positiveWhen I joined Cognition ~1.5 years ago, most devs thought Devin was fake / a scam. Cursor was getting major hype and breaking ARR growth records. Then Claude Code entered the arena, and Devin got pushed even further back in mindshare. It was natural to feel some jealousy and resentment toward the others players, but a… When I joined Cognition ~1.5 years ago, most devs thought Devin was fake / a scam. Cursor was getting major hype and breaking ARR growth records. Then Claude Code entered the arena, and Devin got pushed even further back in mindshare. It was natural to feel some jealousy and resentment toward the others players, but also inspiration from what they were accomplishing. Through it all, @ScottWu46 and @russelljkaplan encouraged us to keep our heads down and keep building against the conviction Cognition has held since day one: cloud agents are the future of software engineering. So far, so good. But the job’s not done, and our heads are still down.view source ↗
-
• twitter 1w positivei am very curious how things work at @openai when it comes to decisions about shipping products, because shipping speed is so fast especially things like, for example, decisions api is @sama just monitoring slack and signs off on things that way? or not involved in approving? i am very curious how things work at @openai when it comes to decisions about shipping products, because shipping speed is so fast especially things like, for example, decisions api is @sama just monitoring slack and signs off on things that way? or not involved in approving?view source ↗
-
• twitter 1w positiveeveryone is building the same harness stack. a loop, memory, mcp adapters, tools, skills, agents.md. then you connect it to slack, linear, notion, your terminal, whatever. these are all indistinguishable clones. so where does the moat actually live? it lives in learned experience. in a vertically fine-tuned stack th… everyone is building the same harness stack. a loop, memory, mcp adapters, tools, skills, agents.md. then you connect it to slack, linear, notion, your terminal, whatever. these are all indistinguishable clones. so where does the moat actually live? it lives in learned experience. in a vertically fine-tuned stack that gets better at one class of problems through thousands or millions of attempts. if you want to build a magical experience for a group of people, say AI for lawyers or AI for biologists, you'll eventually need to own the harness, the experience, and the model. because despite how solved agents can feel right now, models still fail at a lot of simple things. generalization is incomplete. transfer is incomplete. analogy is incomplete. it is very hard to reliably get these things from better prompts and better models. the future will have one very general agent and thousands of vertical agents with their own learned experience, environments, harnesses, and their own models. the harness will appreciate as the experience compounds.view source ↗
-
• twitter 1w positiveAn org chart is partly a map of the expertise that became economical to hire. It leaves out useful capabilities that never justified a whole person. That’s one of the more interesting things AI changes. A company can start maintaining expertise before the work is big enough to become a job. An org chart is partly a map of the expertise that became economical to hire. It leaves out useful capabilities that never justified a whole person. That’s one of the more interesting things AI changes. A company can start maintaining expertise before the work is big enough to become a job.view source ↗
-
• twitter 1w positiveHere is what an average marketing day looks like for me now: 1 minute I'm rebuilding our AEO/SEO workflow. The next I'm in Muse working on Instagram Reels and Meta Ads. Then I'm in Codex improving YouTube thumbnails and headlines. After that, it's ChatGPT Sites, MCPs, plugins, and landing pages. That sounds like a… Here is what an average marketing day looks like for me now: 1 minute I'm rebuilding our AEO/SEO workflow. The next I'm in Muse working on Instagram Reels and Meta Ads. Then I'm in Codex improving YouTube thumbnails and headlines. After that, it's ChatGPT Sites, MCPs, plugins, and landing pages. That sounds like a lot of different jobs at once. But it's also the expectation now. Wednesday I am going live and sharing my screen so you can see exactly how it fits together and take it back to your own team. That is what I want to show you: the way of working behind it, so a small team can get more done without hiring. Wed, Oct 7 11:30 AM PT / 1:30 PM CT Register below! 👇view source ↗
-
• twitter 1w positive200+ creators applied to join the deck 💣 we're stamping cards as fast as we can. founders, your launch crew just got bigger. creators, there's still a spot for you in the deck https://t.co/d9LFdAGRjy 200+ creators applied to join the deck 💣 we're stamping cards as fast as we can. founders, your launch crew just got bigger. creators, there's still a spot for you in the deck https://t.co/d9LFdAGRjyview source ↗
-
• twitter 1w positiveIt’s hard to fathom https://t.co/OL0LzGtvAw getting even faster, but it’s now… much faster 😂 As models speed up (as with Astra 𝚞𝚕𝚝𝚛𝚊𝚏𝚊𝚜𝚝), harness overhead matters more and more. Next release will feature even better performance for session storage and retrieval, and a big unlock for cloud durability in libfx. It’s hard to fathom https://t.co/OL0LzGtvAw getting even faster, but it’s now… much faster 😂 As models speed up (as with Astra 𝚞𝚕𝚝𝚛𝚊𝚏𝚊𝚜𝚝), harness overhead matters more and more. Next release will feature even better performance for session storage and retrieval, and a big unlock for cloud durability in libfx.view source ↗
-
• twitter 1w positivewhat type of prompt or guardrails do people use to effectively run agents overnight, or is it all larp im genuinely curious, feel like im falling behind what type of prompt or guardrails do people use to effectively run agents overnight, or is it all larp im genuinely curious, feel like im falling behindview source ↗
-
• twitter 1w positiveBREAKING: GPT-6 Astra by @OpenAI has taken the #1 spot across 4 of Design Arena’s leaderboards: 3D Design (1484), Frontend (1397), Full Stack (1355), and Image-to-HTML (1272). This is an impressive takeover since its release just a month ago - in particular, it excels at 3D Design (see example below): BREAKING: GPT-6 Astra by @OpenAI has taken the #1 spot across 4 of Design Arena’s leaderboards: 3D Design (1484), Frontend (1397), Full Stack (1355), and Image-to-HTML (1272). This is an impressive takeover since its release just a month ago - in particular, it excels at 3D Design (see example below):view source ↗
-
• twitter 1w positiveBuilder's great opportunity for you. Builder's great opportunity for you.view source ↗
-
• twitter 1w positiveOk hire me Codex app and exec got slopped tooooo much rn Wanna fix Ok hire me Codex app and exec got slopped tooooo much rn Wanna fixview source ↗
-
• twitter 1w positiveGood to start saving at a young age 🫶 https://t.co/YtqmCskSNY Good to start saving at a young age 🫶 https://t.co/YtqmCskSNYview source ↗
-
• twitter 1w positiveThanks for all the Pakistanis who are participating in the long march with all the difficulties created by Form 47 government. May Allah bless you and bless the nation. Thanks for all the Pakistanis who are participating in the long march with all the difficulties created by Form 47 government. May Allah bless you and bless the nation.view source ↗
-
• twitter 1w positiveTop 10 fastest-growing Claude Skill repos of September 2026 (just send this to Claude and tell it to install the most useful ones for you) https://t.co/MArCzQlqfl Top 10 fastest-growing Claude Skill repos of September 2026 (just send this to Claude and tell it to install the most useful ones for you) https://t.co/MArCzQlqflview source ↗