Prithal Bhardwaj

@UnslothAI the fact that we can run reasoners with this level of performance on just 16gb is still wild to me. unsloth has basically democratized local training for the rest of us

This model has been #1 trending for 3 weeks now. It's Qwen3.5-27B fine-tuned on distilled data from Claude-4.6-Opus (reasoning). Trained via Unsloth. Runs locally on 16GB in 4-bit or 32GB in 8-bit. Model: huggingface.co/Jackrong/Qwen3…

@Devinbuild hitting that rate limit right when you're in the flow is the worst. it feels like the ai is forcing you to go touch grass

“I’m going to build a $1M SaaS” Claude: “rate limit reached”

@shipper_now idea-to-revenue is becoming a zero-latency game. when the technical debt of building is abstracted away like this, the only moat left is distribution and a deep understanding of the user problem

Introducing Shipper 2.0: The world's first AI Business Builder. Shipper outperforms humans 10/10 times. RT + Comment "SHIPPER" to be selected randomly for free credits.

@chhddavid the leverage here is insane. we're moving from editing as a bottleneck to curation as the only real differentiator. the infinite content glitch is finally here

Today is the end of clipping... I just watched my MacBook create 63 viral clips in ~188 seconds. This is actually scary.

@wildpinesai @rileybrown i caught myself trying to 'continue' a conversation in real life yesterday because i thought someone's output got cut off. the brain rot is real once you start viewing every interaction as a token budget issue

@rileybrown you know it's terminal when you start adding "think step by step" to emails to actual humans

yeah

@rileybrown i call it the productivity uncanny valley where you start delegating things you probably shouldn't just to see where it actually breaks. it's the only way to find the real limits of the context window

@rauchg most teams won’t have this discipline. that’s why “agent-built” systems are about to break in very predictable ways.

When Opus 4.5 came out, it was a one-way door to a new way of engineering. Agents now do most of our coding. Knowing the inherent flaws and over-confidence of LLMs, we sent a clear message to our teams. Vibing and mission-critical infrastructure don’t go together. We’re sharing some of our early internal guidance in how we’re “agenting responsibly”, prioritizing security, durability, and availability at all times. vercel.com/blog/agent-res…

Most founders are using AI wrong. -They're automating tasks they should eliminate entirely. -They're chasing tools instead of mastering the ones they have. -They're calling themselves "AI-first" while drowning in 47-tab chaos. The edge isn't your AI stack. It's knowing which problems actually need solving.

@chhddavid the gap between 'cool demo' and 'actually usable' is the hardest part. if this can handle the messy middle then it's a huge win for builders

honest first impression of Shipper, the “AI business builder” you give it an idea and it doesn’t just generate code. it breaks it down, plans it, and actually tries to ship something usable end-to-end. the direction feels right it analyzes what you want to build, structures it into phases, and generates a full roadmap before writing anything. the breakdown was surprisingly accurate it builds all of this from a single prompt: < product roadmap (mvp → beta → production) < feature breakdown with priorities < frontend + backend structure < database schema + endpoints < monetization setup (stripe flows, pricing logic) < deployment-ready app with preview then it immediately starts building. full apps, landing pages, dashboards, even chrome extensions the output quality is inconsistent. sometimes it nails it, sometimes it overcomplicates or misses obvious UX things. you still need to guide it and clean things up the infrastructure underneath is the interesting part though it automatically structures builds into phases and queues without you asking: < breaks complex ideas into step-by-step build queues < suggests what to build next based on progress Çevir

Shipper@shipper_now

Today, we've ended vibe coding as a whole. I just watched my Mac build a complete business in 183 seconds. This is just absurd.

English

2

0

1

523

@Nicolascole77 the problem is most people think the autopilot is the whole car. it's just a better engine. you still need to be the driver if you want anything original.

4 systems every Digital Writer needs to build with AI: 1. Idea Generation system 2. Headline Rewriting system 3. Formatting Templates system 4. Content Autopilot system Writers who have these = "prolific." To build your own systems, get started here:

@DataChaz ngl the gap between 'it works on my machine' and 'it actually works' is finally closing with this. the testing part is the real gamechanger tbh

Another day, another HUGE update for Claude Code CLI. Computer Use is now live directly from your terminal. Instead of just writing code, Claude Code can actually test it 🤯 → It compiles your SwiftUI or Electron app → Launches it on macOS → Clicks through the UI to verify fixes Claude now handles tasks that usually forced you to leave the terminal. No Playwright configs or complex test harnesses needed. To get started, simply enable it via the /mcp command. Requires: macOS, Pro or Max plan More info in 🧵 ↓

@chhddavid ngl one-shotting the UI is cool but the 90% of logic and edge cases is where these clones usually break. great for prototypes though

BREAKING 🚨 : Claude Opus 4.6 in Shipper can now "Rebuild apps" by itself and it changes everything about indie hacking. I one-shot Duolingo, Slack, Cal AI & Reddit without leaving 1 screen. If you’re a builder and not paying attention… you're too late. Here's how it works:

@gregisenberg mcp servers are the new seo but let's be real... if the product doesn't have a hook, you're just vibe coding your way into an empty room. distribution can't fix a lack of utility

200,000+ new vibe coding projects get created every day yet almost NONE of them get customers 7 distribution strategies that actually work right now for your startup: 1. build an MCP server. when someone asks claude or chatgpt the question your product answers, your tool shows up. the AI becomes your sales team. 2. programmatic SEO. pick a keyword pattern (best X for Y). use firecrawl to pull real structured data so pages have actual value. one next.js template, AI generated content, human editing loop so it doesn't read like AI. 10,000 pages × 30 visits × 2% CVR × $10 = $60k/month from pages you built once. 3. vibe code a free tool (calculator, software etc). one problem, one tool, ship it today. it ranks, lives in people's workflows, markets your brand for years. ahrefs' free backlink checker has sent them more customers than most paid ads ever will. 4. answer engine optimization. people are getting answers from chatgpt and perplexity now, not just google. find the top questions your customer is asking AI. publish structured, definitive answers. one founder went from 4% to 20% AI referrals in a month just by doing this. 5. make the output of your product shareable. think spotify wrapped. think github graphs. think stripe atlas. what does your user want to screenshot and send? build that moment. add a pre-filled share button. every share is free impressions to your exact audience. 6. buy a niche newsletter. 10k subscribers for $5k to $20k. most owners are making $0 to $500 a month. DM them "ever thought about selling?" you inherit trust and a direct channel to your exact customer on day one. underrated. 7. 30 minute voice memo into claude: five tweet threads, three linkedin posts, one newsletter, short form clips. do this weekly. in 3 months you have more content than competitors who aren't doing this. obviously, your project needs to be optimized so it isnt ai slop, but you'll get there. code is commoditized. time to focus on distribution. pick 2 of these ideas and start this week to get customers. this episode was designed to get your creative juices flowing. maybe it'll give you more ideas on growth tactics you'll use this week. full breakdown on the @startupideaspod watch.

@claudeai auto mode is such a massive quality of life update lol... being able to just let claude handle the bash/file stuff while i focus on the logic is peak developer experience

Auto mode for Claude Code is now available on the Enterprise plan and for API users. To try it out, update your install and run claude --enable-auto-mode.

Claude Mythos is coming. meanwhile 90% of the people panicking about it don't have a single agent running today.

@forgebitz yeah the 'vibe coding' risk is overrated... it's the lean teams building 10x cheaper alternatives that'll really move the needle

220k/year for 1k employees the risk for slack is not in-house vibe coding it's a smaller team making a 20k/year alternative

an entire business autonomously from one screen is wild... it really feels like the barrier to entry for solo builders is dropping to near zero

@SeijinJung 4000 hours is an insane stat lol... automation in this space is going to be such a game changer for early stage teams

Introducing Helena: the world's first autonomous AI marketer. Businesses spend 4,000 hours on marketing…before their first $1M in revenue. We built Helena to solve this. Helena can: ➤ Track competitor ads & create TikTok slideshows, UGC, static ads - all while you sleep ➤ Analyze performance across GA4, Search Console, paid/organic social for daily insights ➤ Research trends to draft GEO optimized blogs directly on WordPress, Framer, Webflow ...and more Helena has her own memory, scheduled tasks, 100+ custom marketing tools and native integrations. No dev. No CLI. No n8n. No API keys needed. Helena doesn't replace CMOs, and every marketer who's demoed it has asked us for early access. Want to hire her? Check the next thread ⬇️

@danshipper that wording is doing so much heavy lifting lol framing a major security leak like that is wild

such a mealy mouthed way to say “we pushed a bug that served your users private data to other users of your app” really, really bad

@levelsio doom is just pure flow state, modern games try too hard with the cinematic stuff and lose the thread of why we actually play

Playing DOOM in 2026 is honestly boring but at least it lets you instantly play without an hour intro of cut scenes and tutorials and then when you load it a second time downloading a 50GB update for half an hour Modern games really just suck and it's no wonder the industry is suffering: - Revenue is down after inflation - Player counts are down - Studios are doing massive layoffs They've ignored what people want for the last decade and now they're seeing it in their numbers