•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 2801–2825 of 4998 mentions
-
• twitter 1w positiveRun Claude and Codex for free. Use a local AI model on your own computer for $0. Then switch back to big cloud models when you need more power. You get: • Free local AI • Private + offline work • Claude + Codex control • Cloud models when needed You don't have to pick just one model. Want the SOP? DM me. Run Claude and Codex for free. Use a local AI model on your own computer for $0. Then switch back to big cloud models when you need more power. You get: • Free local AI • Private + offline work • Claude + Codex control • Cloud models when needed You don't have to pick just one model. Want the SOP? DM me.view source ↗
-
• twitter 1w positiveI have never had so much fun changing the design aspects of my app, like I currently have with Opus 5.5 Medium. It's fast, pretty smart, reliable. Not too overreaching. Understands what I mean in 90% of cases and it ships stable code. This is feeling like magic. @AnthropicAI GG I have never had so much fun changing the design aspects of my app, like I currently have with Opus 5.5 Medium. It's fast, pretty smart, reliable. Not too overreaching. Understands what I mean in 90% of cases and it ships stable code. This is feeling like magic. @AnthropicAI GGview source ↗
-
• twitter 1w positiveHow to spend $300 on ads for $1K in rev < Get Higgsfield < Open Meta ads library < Find goated animated AI and AI UGC ads < Have Claude rewrite them for your apps niche and make a prompt for Higgsfield < Generate in Higgsfield < Absolutely print I’ll do a more in depth tutorial soon but I am happy to admit I w… How to spend $300 on ads for $1K in rev < Get Higgsfield < Open Meta ads library < Find goated animated AI and AI UGC ads < Have Claude rewrite them for your apps niche and make a prompt for Higgsfield < Generate in Higgsfield < Absolutely print I’ll do a more in depth tutorial soon but I am happy to admit I was completely wrong about AI ads. These mfs can convertview source ↗
-
• twitter 1w positiveUsing T3 Code to upgrade T3 Code across my entire fleet at one time feels illegal, but also amazing. 🙃 https://t.co/a11cbbsTkZ Using T3 Code to upgrade T3 Code across my entire fleet at one time feels illegal, but also amazing. 🙃 https://t.co/a11cbbsTkZview source ↗
-
• twitter 1w positiveincredible incredibleview source ↗
-
• twitter 1w positiveYesterday I posted about inference price war pushing margin down. But I since learned inference scammers are a thing with shady price manipulations: As inference providers, - You can lower your input token price 100x cheaper but up your output price 5x. You rank higher in openrouter in price but tots cost is much hi… Yesterday I posted about inference price war pushing margin down. But I since learned inference scammers are a thing with shady price manipulations: As inference providers, - You can lower your input token price 100x cheaper but up your output price 5x. You rank higher in openrouter in price but tots cost is much higher. - You can quantize the model to serve cheaply but users are buying a shittier model. - You can retain and sell user data elsewhere to make more money. - You can secretly swap to DeepSeek v4 flash but say it’s GLM 5.3 10% of the time. Therefore it’s hard to know 1. You get the model you ask for 2. You pay for the tokens you actually burn across cache input, input, output and 3. Your data is safe. I’m sure openrouter is working on it AND baseten, fireworks, together and other enterprise brands will continue to thrive even with higher pricing because they sell peace of mind.view source ↗
-
• twitter 1w positiveHad a bit more time to build out the portfolio page today. Investor sites should look cool feel fast and be informative. Ps @LambLabs ❤️ logo. Oh also, an actually useful search for a portfolio. Have a look, suggestions welcome. https://t.co/Scj8HpDT6z Had a bit more time to build out the portfolio page today. Investor sites should look cool feel fast and be informative. Ps @LambLabs ❤️ logo. Oh also, an actually useful search for a portfolio. Have a look, suggestions welcome. https://t.co/Scj8HpDT6zview source ↗
-
• twitter 1w positiveToday we shipped Memory and Dreaming in Devin: persistent self-cleaning memory for agent sessions. The standard is OSS and also works across any harness in any environment. ⚡ https://t.co/vFscgaI0yN Today we shipped Memory and Dreaming in Devin: persistent self-cleaning memory for agent sessions. The standard is OSS and also works across any harness in any environment. ⚡ https://t.co/vFscgaI0yNview source ↗
-
• twitter 1w positiveWe had the chance to sit down with Arlander Taylor, Ceej Vega, and Mattaniah Aytenfsu at CultureCon, chatting about how they intentionally integrate emerging tools and elevate their craft. They talked about balancing creative boundaries, like how some prefer to protect the early, messy stages of brainstorming, and the… We had the chance to sit down with Arlander Taylor, Ceej Vega, and Mattaniah Aytenfsu at CultureCon, chatting about how they intentionally integrate emerging tools and elevate their craft. They talked about balancing creative boundaries, like how some prefer to protect the early, messy stages of brainstorming, and they also discussed how these tools can be excellent creative partners that help overcome creative blocks. Because at the end of the day, it's about figuring out “how we can leverage [these tools] to amplify the skills that we already have.” Thank you to the CultureCon community for showing up and diving deep with us. Long live the process.view source ↗
-
• twitter 1w positiveAfter 7 years, I’ve made the difficult decision to leave X. Over the years, I’ve helped rebuild the Business and Product from the ground up with a fraction of the original Twitter team, across many significant changes and many long nights of war rooms—scaling many multiples in product revenue, contributing to the early… After 7 years, I’ve made the difficult decision to leave X. Over the years, I’ve helped rebuild the Business and Product from the ground up with a fraction of the original Twitter team, across many significant changes and many long nights of war rooms—scaling many multiples in product revenue, contributing to the early days of xAI, establishing our existing and future partners, supporting the early days of @bot with X, and reinvigorating our Developer Platform as the Everything API for the next generation of agents and developers. Thank you @elonmusk for giving me a chance to thrive as an extreme operator, @jack for hiring me, and our incredible @X engineers and team, who will carry this mission forward as the global town square.🫡❤️view source ↗
-
• twitter 1w positiveI recently started using Grok Bot to make a new website for our machine shop using Wordpress. It’s not live yet but I’m very happy with the whole experience. Website development is one of many industries that will not survive. Companies quoted me 4K-10k to do this I recently started using Grok Bot to make a new website for our machine shop using Wordpress. It’s not live yet but I’m very happy with the whole experience. Website development is one of many industries that will not survive. Companies quoted me 4K-10k to do thisview source ↗
-
• twitter 1w positiveHUGE: We just increased free daily usage by 50% (!) in our coding agent 🎉 Use up to 10 hours of DeepSeek V4.1 Flash, completely free every day (varies by country) Try now: https://t.co/uCHw0PuzKO https://t.co/WygkfnlIsN HUGE: We just increased free daily usage by 50% (!) in our coding agent 🎉 Use up to 10 hours of DeepSeek V4.1 Flash, completely free every day (varies by country) Try now: https://t.co/uCHw0PuzKO https://t.co/WygkfnlIsNview source ↗
-
• twitter 1w positiveAusar Thompson ➡️ John Collins A smooth reverse jam to finish the lob! https://t.co/TRo8xDWYXm Ausar Thompson ➡️ John Collins A smooth reverse jam to finish the lob! https://t.co/TRo8xDWYXmview source ↗
-
• twitter 1w positiveRT @kr0der: this is actually the biggest improvement to my productivity in recent times i didn’t even realise how good it was gonna be i… RT @kr0der: this is actually the biggest improvement to my productivity in recent times i didn’t even realise how good it was gonna be i…view source ↗
-
• twitter 1w positivei know you guys don't know how to read code anymore but it's so nice to be able to just read this https://t.co/0YdXIkqdzy i know you guys don't know how to read code anymore but it's so nice to be able to just read this https://t.co/0YdXIkqdzyview source ↗
-
• twitter 1w positiveIt doesn't make sense to say a subscription plan is worth $ X in isolation. The way to think about subscriptions is that your monthly payment grants you some numbers of “credits”. Each (model, token type) combo consumes a different amount of credits. Because credit cost ratios can differ dramatically from API price ra… It doesn't make sense to say a subscription plan is worth $ X in isolation. The way to think about subscriptions is that your monthly payment grants you some numbers of “credits”. Each (model, token type) combo consumes a different amount of credits. Because credit cost ratios can differ dramatically from API price ratios, the “value” of the same plan changes depending on what model and workload you’re running. Here’s how the API-equivalent value of the same $200/month Claude plan changes depending on the model and workload. (1/6)🧵view source ↗
-
• twitter 1w positivewhy you should care about AI design patterns in your marketing by patterns, I mean the design choices that keep appearing in generated websites, ads and videos the purple gradients, unnecessary labels, cards inside cards and glowing dots moving between boxes you can end up with a finished design that has very little… why you should care about AI design patterns in your marketing by patterns, I mean the design choices that keep appearing in generated websites, ads and videos the purple gradients, unnecessary labels, cards inside cards and glowing dots moving between boxes you can end up with a finished design that has very little connection to your brand your logo and colors might be correct, but the layout, illustrations and motion could belong to any other company that's why recognizing these patterns is useful, especially when you're using agents to produce more of your marketing you need to know which choices fit the work and which ones you're accepting because the model added them here's how to give it better direction: 1. give the agent examples of your approved work include landing pages, ads and videos, with notes on what you want it to follow “use this spacing and typography” gives it more direction than “make it premium” 2. explain what you want to avoid name the specific patterns, like eyebrow text above every heading or pulsing dots along connecting lines a gradient or a card can work, you don't need to ban either from every design 3. review the design before it builds or renders everything ask for a mockup or storyboard first, then review the layout, text and visual choices for motion, agree on what moves, when it moves and what the movement helps explain 4. save your corrections in the brand book keep approved examples alongside the rules, including why you rejected earlier versions give your agents access to this when they create the next asset 5. find useful skills, plugins and harnesses with design instructions look for ones that give your AI specific patterns to avoid and guidance on layout, typography and motion check their examples before using them, then adapt the instructions to your brand studying good design helps you explain what you want, studying AI design patterns helps you recognize what needs changingview source ↗
-
• twitter 1w positive“summarize up to here” in claude code is great bc it lets you inject info from the future above the prefix ie, it *lets you* recompute the embeddings. this is a feature, not a bug, unless you’re a broke ass “summarize up to here” in claude code is great bc it lets you inject info from the future above the prefix ie, it *lets you* recompute the embeddings. this is a feature, not a bug, unless you’re a broke assview source ↗
-
• twitter 1w positiveThis happens to smart people because of: 1. Deep-rooted identity of “I am good at product” (vs. “I am good at winning” or something like that) 2. Moral aversion against marketing & distribution 3. A habit of seeking virtue points by proclaiming “I just want to serve my users” This happens to smart people because of: 1. Deep-rooted identity of “I am good at product” (vs. “I am good at winning” or something like that) 2. Moral aversion against marketing & distribution 3. A habit of seeking virtue points by proclaiming “I just want to serve my users”view source ↗
-
• twitter 1w positiveTraining Qwen-3.8-27B directly on its own harness wins made it WORSE on Terminal-Bench 2 (57% to 53.4% pass@3). Rewrite those same wins as runbooks, re-solve them under a plain harness, then train? 74.2%. Tencent HY + UMD's Recursive Self-Rewrite: 3 harnesses solved 34% more tasks than the best one alone, and 2,001 w… Training Qwen-3.8-27B directly on its own harness wins made it WORSE on Terminal-Bench 2 (57% to 53.4% pass@3). Rewrite those same wins as runbooks, re-solve them under a plain harness, then train? 74.2%. Tencent HY + UMD's Recursive Self-Rewrite: 3 harnesses solved 34% more tasks than the best one alone, and 2,001 wins became 11,094 clean trajectories. Harness tricks are turning into weights. How long before the scaffold is optional? #AgenticAI #PostTraining #AIAgentsview source ↗
-
• twitter 1w positivebasically you’re paying a premium as a startup bc you’re not very creditworthy (also not big enough for the datacenter to get low-interest loans based on you) the solution is ganging up and forming a cooperative, where a bunch of companies all looking for colo space put out a bid together and share the space all the … basically you’re paying a premium as a startup bc you’re not very creditworthy (also not big enough for the datacenter to get low-interest loans based on you) the solution is ganging up and forming a cooperative, where a bunch of companies all looking for colo space put out a bid together and share the space all the companies involved pay less, and the datacenter is happy bc they can bring this to the bank and get a decent loan if you’re interested in this hmu! i’ll put a list togetherview source ↗
-
• twitter 1w positive🗞 Daily AI Brief (2-Min) 🔥GOOGLE : • Gemini app is reshuffling model access: free users drop to Flash-Lite only from Oct 9, AI Plus loses Pro, and AI Pro gains Deep Think. 🔥ALEPH ALPHA : • Kolibri-1 is out as open weights under Apache 2.0: a 78B MoE with 3.46B active parameters, built for German/English sovereign d… 🗞 Daily AI Brief (2-Min) 🔥GOOGLE : • Gemini app is reshuffling model access: free users drop to Flash-Lite only from Oct 9, AI Plus loses Pro, and AI Pro gains Deep Think. 🔥ALEPH ALPHA : • Kolibri-1 is out as open weights under Apache 2.0: a 78B MoE with 3.46B active parameters, built for German/English sovereign deployments. 🔥ANTHROPIC : • Claude now asks voice users to opt in to sharing voice data for model training, via a separate privacy toggle that is off by default. 🔥OPENAI : • GPT-6.1 Sol Ultrafast is coming soon, per Codex lead Tibo Sottiaux. • Safety lead David Robinson resigned, writing in The Atlantic that OpenAI's "culture is broken." 🔥MISTRAL : • Mistral is working on its own Voice Mode, which looks TTS-based for now. 🔥META : • Meta is building an "Agent Engine" for its Meta API Platform, codenamed Forge internally. It'll allow developers to build and run AI agents directly on its own infrastructure.view source ↗
-
• twitter 1w positivebenchmark against known findings glm: 6.5/10 openai: 8/10 claude: 8/10 however glm was like 1/20th cost benchmark against known findings glm: 6.5/10 openai: 8/10 claude: 8/10 however glm was like 1/20th costview source ↗
-
• twitter 1w positivethe original study, free to read (Dell'Acqua, Mollick, Lakhani et al., NBER 2025): https://t.co/fV1FGmRsUL the original study, free to read (Dell'Acqua, Mollick, Lakhani et al., NBER 2025): https://t.co/fV1FGmRsULview source ↗
-
• twitter 1w positiveDidn't know @Muse could ball like that (inspo @arisawyers) https://t.co/ePwU7PqiVn Didn't know @Muse could ball like that (inspo @arisawyers) https://t.co/ePwU7PqiVnview source ↗