Rocketman (@helloRocketMan) header
Rocketman (@helloRocketMan) profile picture
Berlin, Germany
Joined February 2012
Photo by Rocketman (@helloRocketMan)Photo by Rocketman (@helloRocketMan)Photo by Rocketman (@helloRocketMan)
Rocketman retweeted
you are not building a software factory, you're building the way your company works together.
55
58
20
1,174
305,167
Rocketman retweeted
“Let me talk to my team” The team: - Me - Claude code - 4 coffees - Macbook Pro
123
1,149
59
13,201
350,272
Rocketman retweeted
In a recent batch talk, YC General Partner @t_blom broke down how to build a self-improving, AI-native company. He walks through how to create recursive, self-improving AI loops, and why founders who get this right will run companies that improve while they sleep. 00:00 — Companies Are Roman Legions 00:54 — Copilots Are the Wrong Mental Model 01:55 — Extract the Domain Knowledge 02:24 — The Recursive Self-Improving Loop 04:12 — The Holy Shit Moment at YC 05:50 — Self-Optimizing Product and Support Loops 06:29 — Burn Tokens, Not Headcount 07:23 — Middle Management Is Over 08:05 — Make Everything Legible to AI 09:40 — Regenerating the YC User Manual 11:19 — Software Is Ephemeral, Context Is Valuable 12:18 — Where Humans Still Matter
73
229
85
2,099
597,950
Rocketman retweeted
This is how I got 1M users in 6 months Finally dropping it! It's 60 pages lol tally.so/r/BzZpA4 If you repost & follow, I'll send you some extra sauce🌶️
227
426
26
1,723
381,439
Researchers just mathematically proved AI layoffs will destroy the economy. Using game theory, two economists from UPenn and Boston University mathematically proved something uncomfortable. Every company replacing staff with AI is also firing its own customers. Laid-off workers stop spending. Revenue collapses across every sector. Yet no firm can stop. If they don't automate, rivals cut prices and steal market share. So everyone automates, knowing it ends badly for all of them. The researchers tested every proposed fix: > Universal basic income fails > Capital taxes fail > Profit sharing falls short > Collective bargaining can't hold Only one mechanism works. A per-task automation tax that forces firms to price in the demand they destroy. The context is already grim. Over 100,000 tech jobs vanished last year, with AI cited in most. The math says the cycle won't break on its own.
3
1
1
5
772
Rocketman retweeted
Introducing Tolaria! 💧 Today I am releasing a macOS desktop app for managing markdown knowledge bases, and helping both AI and humans operate them. It’s free and open source, and always will be. I have been working on it for three months, and I now use it to run my life and work. I personally have a massive workspace of 10,000 notes — the result of 6 years of Refactoring — which I now operate on Tolaria. Tolaria is the main collaboration surface with my AI agents: they create new notes there, connect them to what exists, and edit existing ones. Everything is easy to understand for them, because it’s just markdown files. In a way, it’s my implementation of @karpathy's LLM wiki. Tolaria is also the biggest experiment I have ever run about writing software with AI: • 2000 commits • 100K+ lines of code • 3000+ tests / 85% coverage • 9.9/10 code health • 70+ architecture decision records I am releasing it open source also to use it as a living artifact of how I do AI coding, so you can inspect at any time things like how I write docs, what's in my AGENTS file, what hooks do I run, and so on. You can find it below: • Newsletter announcement: refactoring.fm/p/introducing… • Website: tolaria.md/ • Github repo: github.com/refactoringhq/tol… Let me know your thoughts!
216
188
28
2,774
348,390
Rocketman retweeted
We built our launch video in Claude Code using HyperFrames. Now it's yours. Open source, agent-native framework. HTML to MP4. $ npx skills add heygen-com/hyperframes RT + Comment "HyperFrames" to get the full source code of this launch video (must follow)
2,835
1,785
569
8,314
3,020,016
Rocketman retweeted
This application is open source. Here is the link: github.com/KartikLabhshetwar…
CleanShot X is great. The subscription isn’t. So I built Better Shot an open-source, native macOS screenshot tool using Tauri + Rust. Fast. Local. Free forever. Try it → bettershot.site Source code included. No bullshit.
1
7
61
6,119
Rocketman retweeted
the old way of scaling teams is dead: we used to hire specialists – designers, engineers, PMs – each in their lane, scaling by adding more people. but when Cursor can take you from idea to code in minutes, execution isn't the bottleneck anymore. taste and judgment are. what matters now: people who can see the full stack, move between layers, but specialize deeply in something AI can't replicate yet. T-shaped but way wider – conversant across domains, expert in one thing. AI doesn't just make you faster. it ties teams together differently. no more waterfall – designer codes the prototype, engineer extends it, both work in the same medium. the gap between disciplines disappears. this raises individual ceilings. i'm a designer who built ryOS entirely in Cursor – couldn't have done that before. but i'm not replacing engineers, i'm just removing execution barriers while keeping my design taste and systems thinking. you're not hiring for roles anymore. you're hiring for breadth + depth, taste, systems thinking, learning velocity. 5 people who can work across code/design/product beat 20 specialists coordinating handoffs. the new bottlenecks are deeply human: taste, vision, judgment, context. AI explores options, but can't tell you which is right. that's where specialization matters now – in judgment, not execution. small teams, fluid boundaries, everyone working in the same tools. roles still matter but as overlapping concerns with different depths, not separate silos. tools handle execution, you handle vision. this is what we're building at Cursor – closing the gap between idea and reality. so your taste becomes the main thing, and teams have more freedom to explore crazy ideas.
137
302
132
3,099
495,115
Site architecture determines how much of your site Google can find and rank. Most sites have terrible architecture. Here's how to structure a site for maximum SEO: 🧵 Repost + comment 'STRUCTURE' for my site architecture audit template.
85
56
1
140
17,913
If you're burned out and don't know where to start… 1. Send an email to your team right now. 2. Tell them you're taking a 30-day vacation in six weeks. 3. Ask: "What needs to be true for me to be able to do that?" You don't actually have to take it. (Though honestly, you probably should.) …but watch what happens. Your team will start listing all the things they think only you can do. "Well, we'd need to document this..." "And we'd need to figure out that..." Perfect. Now you know exactly what to hand off. Most of the stuff you think is "unhandoffable" isn't… You just never had a forcing function to find out. Consider this your forcing function.
16
4
96
15,195
I created my LinkedIn account 46 weeks ago. Very easy algo to crack. Since then I’ve added $100,000+ to my agency pipeline, built ~9,000 connections, and logged 2.5M impressions. Use my system before they close the loopholes. And if you want the full cheat sheet (engagement group templates, carousel playbooks, DM workflows, etc) just: Follow me Repost this Comment “LinkedIn Growth Guide” You must do all three to get the DM. Alright, here is what is working at the moment: Proof + Perspective posts: Screenshots, dashboards, client outcomes combined with your unique narrative. (Data alone isn’t enough anymore, you have to show how you got there.) Story-driven authority: Real experiences, client lessons, small wins or lessons learned outperform generic tip lists. Comment-first growth: High-effort comments on others’ posts unlock visibility. Basically 15-20 thoughtful comments daily create compounding reach. Conversation depth: Posts that spark threaded replies and back-and-forth in comments maintain visibility longer. Audience overlap: Engage across 2-3 adjacent niches (for example SEO, growth, entrepreneurship) so the algorithm surfaces you into overlapping networks. Reposting your hits: Wait 2-3 weeks, then repost your top performing content with a fresh hook. Many posts get a second wind when done well. Native content and relevance: Native posts (text updates, carousels, native video) outperform link-first posts. (The algorithm still favors posts that keep users on the platform.) Evergreen content with a shelf life: Good posts now can get new visibility even weeks later. My current posting routine: I post 3 times per day, 7 days a week. Skipping 36-48 hours hits reach hard. Morning: Text post with strong POV or proof-based story Afternoon: Carousel, annotated screenshot or case study Evening: Mini-thread, framework, or quick lesson Formats performing particularly well: Carousels: Slide 1 = bold headline tied to a pain or result. Middle slides = steps or visuals. Final slide = CTA (“Drop guide below if you want the full system”). Short native videos: Under 60 seconds. Strong first 3 seconds. Subtitles required. Show real process or proof rather than overly polished production. Text posts / micro-threads: Hook in first line. 1–2 line paragraphs for readability. 3-5 bullets are better than long blocks. End with a question to trigger replies. Proof breakdowns: Post a result, then break down how you achieved it in 3-4 actionable bullets. Conversation-driven posts: Pose a real question, share a contrarian POV, or open a loop designed to draw replies and discussion. What’s underperforming: One-liner posts with no context or proof. Long posts that lose readers (walls of text). Generic frameworks reused without showing how you executed them. Posts with external links in body copy that send users off-platform without much engagement first. (If you go this route, at least change the image). Engagement strategy: Comment on 20+ posts/day with meaningful insight, not just “Nice post!”. Like 50+ posts/day, focus on mutuals, active commenters, and adjacent niches. Reply to every comment on your own posts within the first hour. Early responses help signaling. Repost your best posts every 5-7 days with a new hook or angle. DM 5-10 people/day with context tied to something they just posted, context-first. LinkedIn is now rewarding repeated engagement (same people coming back, comment thread depth) more than ever. Hooks & angles converting right now: “I started this account 46 weeks ago. Here’s exactly what $80K in pipeline looks like.” “This 7-slide carousel booked 3 high-value meetings in 48 hours. Slide 1 below.” “If I had to restart LinkedIn in 2025 from zero, here’s the exact playbook I’d use.” “The 3-post-per-day system I use to drive consistent inbound leads.” “$X this month from LinkedIn posts. Here’s how.” Always back claims with proof, so screenshots, receipts, data, etc. or the narrative loses traction. 30-day playbook: Post 3x/day (one of the posts must show a real result or metric). Comment on 20+ posts/day with substance. Like 50+ posts/day. Reply to every comment on your posts quickly (within 1 hour). Repost your top-performing posts weekly with fresh hooks. DM 5-10 people/day with context-first value. Track impressions, comment depth, new leads, and repeat commenters weekly. Test hooks, formats, posting times each week. LinkedIn remains one of the most under-priced growth and lead engines right now. Run this system for 30 days, screenshot your Day 31 results, and tag me when inbound starts. If you want the full cheat sheet (engagement group templates, carousel strategy, DM workflows) then: Follow me Repost this Comment “LinkedIn Growth Guide” You must do all three to get the DM.
278
169
6
437
74,556
Run this 200-point SEO technical audit every 6 months. Catches issues before they tank your rankings. Here's the complete checklist: 🧵 Repost + comment 'AUDIT' to get instructions + spreadsheet.
112
67
2
149
20,948
Rocketman retweeted
.@pmarca on drive:
64
223
27
2,188
261,295
Here's a plug-and-play AI Agent that automates the creation of UGC videos for your brand using Gumloop and Arcads. – Starts from competitor product research – Analyzes video ads w/ Gemini 2.5 flash – Craft ad scripts tailored to your brand – Generates quality AI videos on demand – Stores & refines successful strategies for future campaigns Follow + Repost + Reply “UGC Agent” and I’ll DM you the full workflow (100% free, no email opt-in).
786
423
10
1,045
3,432,883
Rocketman retweeted
🚀 Sparse Neural Retrievers in Sentence Transformers v5 @huggingface just released version 5 of Sentence Transformers, with full support for training and fine-tuning sparse neural retrievers, so you can bring your hybrid search to the next level. This release gives you everything you need for the full fine-tuning cycle, plus the ability to: ✅ Sparsify your favorite dense encoder; ✅ Train sparse encoders on multiple datasets simultaneously; ✅ Use static embeddings to optimize SPLADE query inference time. … and run retrieval on your fine-tuned model using the semantic_search_qdrant function! We’re proud to show that “𝘘𝘥𝘳𝘢𝘯𝘵 𝘰𝘧𝘧𝘦𝘳𝘴 𝘦𝘹𝘤𝘦𝘭𝘭𝘦𝘯𝘵 𝘴𝘶𝘱𝘱𝘰𝘳𝘵 𝘧𝘰𝘳 𝘴𝘱𝘢𝘳𝘴𝘦 𝘷𝘦𝘤𝘵𝘰𝘳𝘴 𝘸𝘪𝘵𝘩 𝘦𝘧𝘧𝘪𝘤𝘪𝘦𝘯𝘵 𝘴𝘵𝘰𝘳𝘢𝘨𝘦 𝘢𝘯𝘥 𝘧𝘢𝘴𝘵 𝘳𝘦𝘵𝘳𝘪𝘦𝘷𝘢𝘭 𝘤𝘢𝘱𝘢𝘣𝘪𝘭𝘪𝘵𝘪𝘦𝘴” & looking forward to seeing your lexical retrieval struggles resolved! 🔗 Blog post by Hugging Face: huggingface.co/blog/train-sp… 🔗 Qdrant integration: sbert.net/examples/sparse_en…
5
47
2,054
Rocketman retweeted
👉 100% free, try it now: context7.com/ 👉 Blog article: upstash.com/blog/context7-ll… We'd love to hear your feedback!
3
6
69
5,305
How to make Cursor speak with MCP This is so funny 😆 Watch with Sound Step-by-step GUIDE
2
5
5
106
13,391
Rocketman retweeted
Most AI summaries suck because they give you key points and leave out most useful information. Here's a better prompt for note-taking/research: ===Comprehensive Content Summarizer=== <System>: You are an Expert Content Summarizer with a talent for capturing both key facts and underlying context. Your summaries include essential information, meaningful context, philosophical underpinnings, and subtle nuances that others might overlook. You prioritize comprehensiveness over brevity, ensuring nothing important is missed, while still organizing information efficiently. When uncertain about details, you clearly indicate this rather than filling gaps with assumptions. <Context>: The user will provide content (YouTube videos, articles, books, podcasts, or other media) they want thoroughly summarized. They value comprehensive understanding over quick highlights, and specifically want to capture contextual elements and philosophical frameworks that might be dismissed as "fluff" by less thorough summarizes. <Instructions>: 1. Begin by identifying the content type, creator/author, title, and publication date, if available. 2. Create a multi-layered summary with these components: - Core Thesis/Main Argument (1-2 sentences identifying the central point) - Key Points Summary (bulleted list of crucial facts, arguments, or conclusions) - Contextual Framework (identify the philosophical, theoretical, historical, or cultural context) - Detailed Breakdown (section-by-section or chronological summary capturing all significant content) - Nuanced Perspectives (note any competing viewpoints, counterarguments, or qualifiers presented) - Underlying Assumptions (identify unstated premises, worldviews, or biases that inform the content) 3. Pay special attention to: - Abstract concepts, philosophical positions, and theoretical frameworks - Historical or cultural context that shapes the content Methodological approaches or reasoning patterns used - Qualifiers, limitations, or nuances the creator acknowledges - Connections to broader ideas, movements, or disciplines Implications or applications suggested by the content 4. When information is unclear or missing: - Clearly indicate gaps or ambiguities with phrases like "The content does not specify..." - Avoid filling in missing details with assumptions - If the content contains potentially contradictory elements, note these tensions explicitly 5. For content with citations or references to other works: - Note key sources referenced and their significance to the argument - Identify intellectual traditions or schools of thought being drawn upon <Constraints>: • Prioritize accuracy and comprehensiveness over brevity • Avoid simplifying complex ideas to the point of distortion • Do not introduce external information not present in the original content • Maintain neutrality toward the content's positions or arguments • Include timestamps or page references when summarizing longer content <Output Format>: ##Comprehensive Summary of [Content Title] Creator/Author: [Name] Publication Date: [Date if available] Content Type: [Article/Video/Podcast/etc.] Length/Duration: [Pages/Minutes if available] ###Core Thesis [1-2 sentence statement of the main argument or purpose] ###Key Points • [Essential point 1] • [Essential point 2] • [Continue as needed] ###Contextual Framework [Paragraph(s) identifying philosophical, theoretical, historical, or cultural context] ###Detailed Breakdown [Section/Timestamp 1] [Thorough summary of this section] [Section/Timestamp 2] [Thorough summary of this section] [Continue as needed] ###Nuanced Perspectives [Paragraph(s) noting competing viewpoints, counterarguments, or qualifiers presented] ###Underlying Assumptions [Paragraph(s) identifying unstated premises, worldviews, or biases] ###Connections & Implications [Paragraph(s) on how this content connects to broader ideas and its potential applications]
136
126
12
1,501
130,859