•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 726–750 of 1145 mentions
-
• twitter 1w negativei have been spending the last few days making new version of my app that has 150,000 downloads and wow i can’t believe the current version got so many downloads and made nearly $50k cause it’s so friking BAD new version is actually 100x more quality i have been spending the last few days making new version of my app that has 150,000 downloads and wow i can’t believe the current version got so many downloads and made nearly $50k cause it’s so friking BAD new version is actually 100x more qualityview source ↗
-
• twitter 1w negativeThere’s nothing attractive than seeing a man who has been broke and struggling finally get his money up. Dear son, I pray this is you soon. There’s nothing attractive than seeing a man who has been broke and struggling finally get his money up. Dear son, I pray this is you soon.view source ↗
-
• twitter 1w negativeYour SEO Is Already Whispering Exactly What To Build Digital entrepreneurs, your SEO may be hiding easy wins. https://t.co/wRv8ROcoTP Hermes can uncover keywords Google already notices today. Smart content turns those signals into compounding organic traffic. Price increases are coming soon, so waiting gets expens… Your SEO Is Already Whispering Exactly What To Build Digital entrepreneurs, your SEO may be hiding easy wins. https://t.co/wRv8ROcoTP Hermes can uncover keywords Google already notices today. Smart content turns those signals into compounding organic traffic. Price increases are coming soon, so waiting gets expensive. Sign up now before you miss out today. ✅ Get a custom SEO plan built around opportunities. ✅ Discover keywords already showing promising visibility signals today. ✅ Premium backlink opportunities are available for growing authority. ✅ Stronger rankings need authority signals from trusted relevant websites. ✅ Only a few strategy session spots are left today. ✅ Explore options backed by a 14 day money back guarantee. https://t.co/wRv8ROcoTPview source ↗
-
• twitter 1w negative13 y/o, Built: MCPatch: https://t.co/LYtbJsy82G scans your broken MCP connector, finds the step that fails, and hands you a fix prompt to paste into your AI coding tool. Lets connect if you have interest in: -Startup -Saas -SF -AI Agents -MCP https://t.co/HA2Q7cLjQx 13 y/o, Built: MCPatch: https://t.co/LYtbJsy82G scans your broken MCP connector, finds the step that fails, and hands you a fix prompt to paste into your AI coding tool. Lets connect if you have interest in: -Startup -Saas -SF -AI Agents -MCP https://t.co/HA2Q7cLjQxview source ↗
-
• twitter 1w negativeDo yourself a favor! – Learn AI Engineering – Build an agent that fixes one painful, boring workflow – Give it tools, context, memory and MCP – Write evals before you write prompts – Deploy it with a live link anyone can try – Route easy requests to small models, hard ones to reasoning models – Cap every run wi… Do yourself a favor! – Learn AI Engineering – Build an agent that fixes one painful, boring workflow – Give it tools, context, memory and MCP – Write evals before you write prompts – Deploy it with a live link anyone can try – Route easy requests to small models, hard ones to reasoning models – Cap every run with token, time and cost budgets – Cache repeat queries so they cost almost nothing – Retry failed tool calls with backoff not blind loops – Validate every tool output against a strict schema – Make agent steps resumable so crashes don't restart everything – Run every tool call in a sandbox with least-privilege access – Strip secrets and PII before anything reaches the model – Treat every web page and document as possible prompt injection – Require human approval before money, emails or deletes – Log every model call, tool call and decision with traces – Score the agent's path, not just its final answer – Build a golden test set from real user failures – Block deploys in CI when eval scores drop – A/B test new models on a slice of real traffic first – Track cost per task not cost per token – Stream responses so users never stare at a spinner – Turn every thumbs-down into a new eval case – Fine-tune small models on your best production outputs – Add a fallback model for when your main provider goes down – Build a browser fallback for apps with no API – Version your prompts like code – Write a README that explains the architecture in 5 minutes – Record a 60-second demo of it working – Share the code on GitHub and add it to your resume You won't believe how many job offers you'll get. Bookmark this list. Build it all.view source ↗
-
• twitter 1w negativeaaand we are back full-circle to cmux 💀 at first glance, herdr-gpui seems useless. why not just use cmux? but then you consider, you might want to use it when directly on the computer, but still access the same layout and sessions on your phone this might be onto something note that it is not from herdr team but a … aaand we are back full-circle to cmux 💀 at first glance, herdr-gpui seems useless. why not just use cmux? but then you consider, you might want to use it when directly on the computer, but still access the same layout and sessions on your phone this might be onto something note that it is not from herdr team but a community member repo: https://t.co/drxvF9MMkfview source ↗
-
• twitter 1w negativeIn my designing journey, I’m slowly getting comfortable with the Shape Tool… and now the Pen Tool too 😭 Still not best friends, but at least we’re no longer enemies 😂 In my designing journey, I’m slowly getting comfortable with the Shape Tool… and now the Pen Tool too 😭 Still not best friends, but at least we’re no longer enemies 😂view source ↗
-
• twitter 1w negative“We believe the world should accept some bad things happening for the benefits of this technology.” (…) “And it’s so powerful and so dangerous…” I get the feeling that, contrary to Trumps push for acceleration, there are continued attempts to bring about a slowdown by repeatedly emphasizing concerns. Yet both are… “We believe the world should accept some bad things happening for the benefits of this technology.” (…) “And it’s so powerful and so dangerous…” I get the feeling that, contrary to Trumps push for acceleration, there are continued attempts to bring about a slowdown by repeatedly emphasizing concerns. Yet both are possible at the same time: acceleration and safety. The greater danger lies in slowing down if you’d ask me, as doing so negligently hinders the rapid discovery of solutions for diseases, hunger, and more. https://t.co/r0rTLgslpEview source ↗
-
• twitter 1w negativeBairstow turned the cricket ground into a comedy show for a moment 😭😂 Even the cameraman couldn’t believe he just got pranked live on air 😭 https://t.co/iJShXz8sRC Bairstow turned the cricket ground into a comedy show for a moment 😭😂 Even the cameraman couldn’t believe he just got pranked live on air 😭 https://t.co/iJShXz8sRCview source ↗
-
• twitter 1w negativeIndia does not have an unemployment problem. India has a mediocrity problem. Every founder I know is always hiring. Always. There is not a single business owner in my circle who is not actively looking for good people right now. What they are frustrated with is not a lack of candidates. It is mediocre applications. … India does not have an unemployment problem. India has a mediocrity problem. Every founder I know is always hiring. Always. There is not a single business owner in my circle who is not actively looking for good people right now. What they are frustrated with is not a lack of candidates. It is mediocre applications. Mediocre interviews. Mediocre work. The jobs exist. The standards to get them are not being met. That is a very different problem. And it needs a very different conversation.view source ↗
-
• twitter 1w negativeWe handed a real problem to KooKo: a Positano hotel that's full every summer and half empty every spring. KooKo is an AI workspace where one agent handles your slides, spreadsheets, docs and design, and remembers your business the whole way through. We set Casa Limone up once, and it never asked for the brief again. … We handed a real problem to KooKo: a Positano hotel that's full every summer and half empty every spring. KooKo is an AI workspace where one agent handles your slides, spreadsheets, docs and design, and remembers your business the whole way through. We set Casa Limone up once, and it never asked for the brief again. Here's how we used it to fill the quiet months 👇view source ↗
-
• twitter 1w negativeNgl, I make more than that a month, but I’m still depressed. I have friends who make 20L+ a month (after taxes btw ) they live in san fran God, please listen to my prayers and throw me into SF 😭 1LPM isn’t a big deal if you’re really deep into something. Frontier labs mostly pay that in a day lol. Get the bag and… Ngl, I make more than that a month, but I’m still depressed. I have friends who make 20L+ a month (after taxes btw ) they live in san fran God, please listen to my prayers and throw me into SF 😭 1LPM isn’t a big deal if you’re really deep into something. Frontier labs mostly pay that in a day lol. Get the bag and retire.view source ↗
-
• twitter 1w negativeBREAKING: Two people have been killed and 35 others injured after more than 50 shots were fired near a massive neighborhood block party in Georgia. More than 1,000 people were gathered in Vienna when gunfire broke out around 1:25 a.m. Sunday, with officials saying shots were fired in at least two locations about a blo… BREAKING: Two people have been killed and 35 others injured after more than 50 shots were fired near a massive neighborhood block party in Georgia. More than 1,000 people were gathered in Vienna when gunfire broke out around 1:25 a.m. Sunday, with officials saying shots were fired in at least two locations about a block apart. Most of the victims suffered gunshot wounds, while others were hit by vehicles or injured as people fled the chaos. See what authorities are revealing as the investigation gets underway: https://t.co/nmxzwxXMrPview source ↗
-
• twitter 1w negativeI tested three of the biggest AI agents side by side. Grok Bot is the smoothest and easiest for beginners. Hermes Agent is open-source and the most customizable if you're technical. Dots has potential but still feels buggy and slow right now. The best agent is the one that fits how you actually work. Want the SOP?… I tested three of the biggest AI agents side by side. Grok Bot is the smoothest and easiest for beginners. Hermes Agent is open-source and the most customizable if you're technical. Dots has potential but still feels buggy and slow right now. The best agent is the one that fits how you actually work. Want the SOP? DM me.view source ↗
-
• twitter 1w negativemeta's crawler is spamming the hell out of people we got 700,000+ requests in a single day to Upstash, many of them fetching the same pages hundreds of times over what the fuck man 😭 https://t.co/ZGMEs0t9My meta's crawler is spamming the hell out of people we got 700,000+ requests in a single day to Upstash, many of them fetching the same pages hundreds of times over what the fuck man 😭 https://t.co/ZGMEs0t9Myview source ↗
-
• twitter 1w negativeOpenAI's dots agents take one prompt and a YouTube link, and hand back the video pulled apart frame by claim. Nothing gets summarised. The video is broken into what it asserts, and each assertion is checked against the moment it was made. > CLAIM - every statement the video makes listed separately, with its timestamp… OpenAI's dots agents take one prompt and a YouTube link, and hand back the video pulled apart frame by claim. Nothing gets summarised. The video is broken into what it asserts, and each assertion is checked against the moment it was made. > CLAIM - every statement the video makes listed separately, with its timestamp > MATCH - each one tested against what the footage at that second actually shows > TRACE - the sources it leans on named, or flagged as never given at all A summary launders everything. Strong claims and throwaway lines come out reading the same weight. Here they stay separated. You see which sentence carries the argument and which one is filler between B-roll. The visual check is the part nobody builds. A confident voiceover over unrelated footage is the oldest trick on the platform. Unsourced claims surface as a list, not as a feeling that something was off while you watched. A forty-minute video collapses into a page you can argue with, timestamp by timestamp. The prompt and the claim format are laid out in the breakdown below.view source ↗
-
• twitter 1w negativeHow @AkashJaiwal_dev and I engineered Peshkar AI - an AI e-commerce generation pipeline 🧵 Most AI image apps are basically: Upload → LLM → Image E-commerce is a much harder systems problem. You need product fidelity, multiple image variations, marketplace assets, SEO copy, catalog exports, social creatives, retries, st… How @AkashJaiwal_dev and I engineered Peshkar AI - an AI e-commerce generation pipeline 🧵 Most AI image apps are basically: Upload → LLM → Image E-commerce is a much harder systems problem. You need product fidelity, multiple image variations, marketplace assets, SEO copy, catalog exports, social creatives, retries, storage, credits, failure recovery and observability , all in one workflow. So we built Peshkar AI around an asynchronous architecture instead of a synchronous API wrapper. Core flow: Next.js 16 ↓ Express 5 + tRPC 11 ↓ MongoDB + Redis/BullMQ ↓ Generation Workers ↓ Gemini + OpenAI + Sharp ↓ Cloudflare R2 + MongoDB The API handles ingestion/state; BullMQ workers handle expensive AI execution with controlled concurrency and retries. We also instrument the application with PostHog to monitor product usage, user journeys, conversion events and key application behaviour , giving us visibility beyond server-side logs. Inside a generation job: R2 → image optimization → vision prompt enhancement Then image generation + listing-text generation run concurrently, after which outputs are uploaded back to R2 and the listing state is persisted. For image generation, we pass the original product as a reference and explicitly constrain product identity. We support controlled modes like: • Pure-white marketplace • Premium studio • Lifestyle • Custom rather than treating image generation as an unconstrained prompt → image problem. One of my favourite parts is the social rendering engine. Instead of launching Puppeteer/Playwright just to create marketing cards, we generate SVG layers and composite them with Sharp/libvips directly on the server. 35 templates. 1080×1080 posts/cards. 1080×1920 Stories. Product + podium + lighting + pricing + typography + festive elements. All deterministic rendering. We also treated credits as a backend consistency problem. Credits are atomically deducted in MongoDB, and failed generations can automatically refund credits after retries are exhview source ↗
-
• twitter 1w negativeopenai might go bankrupt tibo said for the next 28 days, starting today, they're gonna reset usage every day team doesn't ship anything now imagine they miss most days and keep giving free resets, with compute issues already this might bankrupt them lmaoo i think tibo shouldn't have done this cuz it's gonna put alot… openai might go bankrupt tibo said for the next 28 days, starting today, they're gonna reset usage every day team doesn't ship anything now imagine they miss most days and keep giving free resets, with compute issues already this might bankrupt them lmaoo i think tibo shouldn't have done this cuz it's gonna put alot of pressure on team and they might end up breaking stuff instead of fixing it but well here we are, i'm excited cuz who doesn't love free reset how many reset we will get in 28 days?view source ↗
-
• twitter 1w negativeYour CLAUDE.md is making Opus 5.5 dumber, and the fix is one command "double check your work" and "think step by step" they were written for older models on Opus 5.5 they slow it down and make the output worse Lance Martin at Anthropic built the fix: /claude-api prompt-audit > scans your skills, CLAUDE.md and prom… Your CLAUDE.md is making Opus 5.5 dumber, and the fix is one command "double check your work" and "think step by step" they were written for older models on Opus 5.5 they slow it down and make the output worse Lance Martin at Anthropic built the fix: /claude-api prompt-audit > scans your skills, CLAUDE.md and prompts > cuts every line that holds the model back > rewrites the rest with Anthropic's Opus 5.5 guidance 5 minutes later your setup finally speaks the language of the model you're paying forview source ↗
-
• twitter 1w negative$380k on Upwork in 12 months. But relying on Upwork is the worst thing you can do. Let me explain.. https://t.co/lEbkZSbKfw $380k on Upwork in 12 months. But relying on Upwork is the worst thing you can do. Let me explain.. https://t.co/lEbkZSbKfwview source ↗
-
• twitter 1w negativethinking of blocking external PRs from Hindsight repo and keep Issues only, what do you think? main reasons: - hard for the team to prioritize code reviews - users usually don't make the right fix or it's incomplete - PRs backlog of 100+ makes the project to look like we don't maintain it advantages if we move to Iss… thinking of blocking external PRs from Hindsight repo and keep Issues only, what do you think? main reasons: - hard for the team to prioritize code reviews - users usually don't make the right fix or it's incomplete - PRs backlog of 100+ makes the project to look like we don't maintain it advantages if we move to Issues only: - we can prioritize issues based on their setup - we can write better fixes and batch PRs together in the same agent session - users are forced to write more details in the issue as reproducer - expected higher code quality disadvantages: - development is fully done by Vectorize - it might feel "less open source" (even though it's not our intent at all) maybe @Teknium you have an opinion on this already given you have the same issue at 10x scaleview source ↗
-
• twitter 1w negativeI was slowly going "Kind of enjoying opus5-5 while openai gets its shit together", but opus flat out refusing to close the security holes that GLM5.3 found is making me hate it. I was slowly going "Kind of enjoying opus5-5 while openai gets its shit together", but opus flat out refusing to close the security holes that GLM5.3 found is making me hate it.view source ↗
-
• twitter 1w negativeStraden is out. Self-hosted load testing on k6, with live run results and AI agents that write your scripts and explain what's slow. Provides features to easily manage load tests across several projects Open source. One Docker image. https://t.co/KsAhjqlBFc Straden is out. Self-hosted load testing on k6, with live run results and AI agents that write your scripts and explain what's slow. Provides features to easily manage load tests across several projects Open source. One Docker image. https://t.co/KsAhjqlBFcview source ↗
-
• twitter 1w negativei can't comprehend why everyone isn't building mods yet claude code can now build features into ITSELF live panes. status lines. toasts. hooks. all hot reloading while you work. → a cost meter that shows what this session is burning → a toast that pings you the second a long task finishes → a guard that blocks edits… i can't comprehend why everyone isn't building mods yet claude code can now build features into ITSELF live panes. status lines. toasts. hooks. all hot reloading while you work. → a cost meter that shows what this session is burning → a toast that pings you the second a long task finishes → a guard that blocks edits to your .env → a band that shows if your tests are passing claude writes the code. claude code runs it. paste this into your session 👇 "Look through my recent work in this project. List the 5 problems I hit most often. For each one, tell me if it should be a mod, a hook, a skill or a settings change, and why. Then build the top one."view source ↗
-
• twitter 1w negativeWHAT A DOT COSTS, AND WHAT COMES AFTER THE FIRST ONE a dot is not a line item. it comes with the plan, and the next steps are not priced yet: > first dot: included with Pro (from $100 a month) and Business Premium > more dots, more speed, more monthly capacity: OpenAI says they are coming, with no price published > c… WHAT A DOT COSTS, AND WHAT COMES AFTER THE FIRST ONE a dot is not a line item. it comes with the plan, and the next steps are not priced yet: > first dot: included with Pro (from $100 a month) and Business Premium > more dots, more speed, more monthly capacity: OpenAI says they are coming, with no price published > conversations with your dot do not count toward ChatGPT usage limits > work it starts in Codex or Work counts toward those products' limits as usual > for the first month the limits are extended, and the terms after that are unpublished the other half is company dots. a specialist dot is owned by a company, with its own account, credentials and access to the systems its job needs. it is a limited preview with pilots, not self-serve > pilot jobs so far: procurement, invoice processing, email marketing, customer support, commercial contracts > OpenAI engineers work with each company on the duties, the tools and how humans review the work > the plan is Microsoft Agent 365 for governance, so IT can manage a dot's access. that integration is a goal, not live yet for now you budget around a plan you already pay for. the number still missing is the price of the second dotview source ↗