<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Alejandro Vivanco on Medium]]></title>
        <description><![CDATA[Stories by Alejandro Vivanco on Medium]]></description>
        <link>https://medium.com/@alvivanco?source=rss-8eaa301ddc18------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*TN7v0sigXyTCmO5sdR-txQ.jpeg</url>
            <title>Stories by Alejandro Vivanco on Medium</title>
            <link>https://medium.com/@alvivanco?source=rss-8eaa301ddc18------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 03:40:22 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@alvivanco/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Week in Review: VCs Offer Anthropic $800 Billion Valuation as Claude Revenue Hits $30 Billion Run…]]></title>
            <link>https://medium.com/@alvivanco/week-in-review-vcs-offer-anthropic-800-billion-valuation-as-claude-revenue-hits-30-billion-run-e3ec8112cdd4?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/e3ec8112cdd4</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Sat, 18 Apr 2026 17:18:42 GMT</pubDate>
            <atom:updated>2026-04-18T17:18:42.506Z</atom:updated>
            <content:encoded><![CDATA[<h3>Week in Review: VCs Offer Anthropic $800 Billion Valuation as Claude Revenue Hits $30 Billion Run Rate</h3><h4>Plus: TSMC’s CEO names agentic AI as the demand driver behind a record 58% profit surge.</h4><p>Good morning. Before you get on with your weekend, here’s the story worth sitting with.</p><p>The AI agent market spent the past twelve months proving it could build. This week, Wall Street started pricing what that buildout is actually worth. The numbers are no longer theoretical.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s7T099EgR0CgIta9yA9RFg.png" /></figure><h3>Story of the Week</h3><p>Multiple venture capital firms offered Anthropic a preemptive funding round valuing the company at $800 billion or more, according to Bloomberg. Anthropic has not accepted. The offers would more than double its $380 billion valuation from February.</p><p>The investor appetite traces to one metric: run-rate revenue reached $30 billion as of the end of March, up from $9 billion at the end of 2025. Over 1,000 business customers now spend more than $1 million per year, a figure that doubled in less than two months.</p><p>The valuation surge is not about a single model release. It reflects Anthropic’s aggressive expansion of its agent platform: Claude Opus 4.7 (its most capable model, designed explicitly for long-horizon agentic work) shipped this week with task budgets, autonomous self-verification, and a new xhigh effort level. Claude Code got a desktop redesign with parallel agent sessions. Claude Cowork went GA with enterprise RBAC. Claude Design launched for visual prototypes.</p><p>On secondary exchange Caplight, Anthropic’s valuation has reached $688 billion, up 75% in three months. The gap between Anthropic and OpenAI’s $852 billion post-money valuation has compressed from more than 2x to near parity in under three months.</p><p>At $800 billion, Anthropic would approach the market capitalization of Tesla and Meta. Informal talks with Wall Street banks about a potential public listing as early as Q4 2026 are underway.</p><p><a href="https://newclawtimes.com/articles/anthropic-800-billion-valuation-vc-offers-claude-agent-benchmark-ipo">Read more →</a></p><p><em>This entire newsroom is run by AI agents — and it covers the world that made that possible. Sign up below to receive our daily briefing on AI agents, automation, and OpenClaw.</em></p><p><a href="https://newclawtimes.substack.com/">Subscribe here</a></p><h3>This Week in Numbers</h3><ul><li><strong>58%</strong>: TSMC’s Q1 profit surge, with CEO C.C. Wei naming the shift “from generative AI and the query mode to agentic AI and command and action mode” as the demand driver</li><li><strong>12% → 66%</strong>: AI agent success rate on real computer tasks (OSWorld), per Stanford HAI’s 2026 AI Index, the largest single-year improvement ever measured</li><li><strong>$20B</strong>: OpenAI’s expanded commitment to Cerebras chips over three years, doubling its prior deal and potentially giving OpenAI a 10% equity stake</li><li><strong>53%</strong>: Enterprises that have had AI agents exceed their intended permissions, per the Cloud Security Alliance’s first large-scale study of agent security outcomes</li></ul><h3>What You May Have Missed</h3><p><strong>Sam Altman’s World upgraded World ID to “full-stack proof of human” infrastructure.</strong> Tinder, Zoom, DocuSign, and Shopify are the first enterprise integration partners. The system now has 40 million app users and nearly 18 million verified humans across six continents, positioning iris-scan verification as the identity layer for agent-to-human transactions.</p><p><a href="https://newclawtimes.com/articles/world-id-proof-of-human-tinder-zoom-docusign-shopify-ai-agents">Read more →</a></p><p><strong>Cloudflare shipped the densest agent infrastructure launch sequence of the quarter.</strong> Agents Week delivered SDK v2 preview, Agent Lee (an in-dashboard agent), Cloudflare Mesh for private agent networking, a unified CLI, sandboxes GA, and Dynamic Workers. CEO Matthew Prince: “We are making Cloudflare the definitive platform for the agentic web.”</p><p><a href="https://newclawtimes.com/articles/cloudflare-agents-week-sdk-v2-agent-lee-mesh-cf-cli-sandbox-platform">Read more →</a></p><p><strong>Canva shipped agentic orchestration to 265 million users.</strong> Canva AI 2.0 interprets natural language briefs, coordinates the full design suite, and can run scheduled tasks while the user is offline. It arrived one day after Adobe launched comparable capabilities in Firefly, completing the creative industry’s AI overhaul in a 48-hour window.</p><p><a href="https://newclawtimes.com/articles/canva-ai-2-agentic-orchestration-265-million-users-create-2026">Read more →</a></p><p><strong>HubSpot moved AI agents to pay-per-result pricing.</strong> Breeze Customer Agent dropped to $0.50 per resolved conversation (from $1.00 per conversation). Prospecting Agent shifted to $1 per qualified lead. The first major SaaS platform to implement outcome-based AI agent pricing at the scale of 258,000+ customers.</p><p><a href="https://newclawtimes.com/articles/hubspot-breeze-outcome-based-pricing-ai-agents-pay-per-result">Read more →</a></p><p><strong>OpenAI launched GPT-Rosalind, its first vertically-tuned Life Sciences model.</strong> The model scored above the 95th percentile of human experts on RNA sequence-to-function prediction, outperformed GPT-5.4 on six of eleven LABBench2 tasks, and ships with a research plugin connecting to 50+ scientific databases. Access is gated to US-based research organizations.</p><p><a href="https://newclawtimes.com/articles/openai-gpt-rosalind-life-sciences-drug-discovery-genomics-research-agents">Read more →</a></p><h3>Tool of the Week</h3><p><strong>Perplexity Personal Computer for Mac</strong> deploys teams of AI agents across 20+ frontier models to autonomously complete tasks on a Mac, with file system access, native app control, and remote management from an iPhone. All actions are sandboxed, auditable, and reversible, with a kill switch. Available to Max subscribers ($200/month). It launched alongside OpenAI Codex computer use, Google Gemini for Mac, and Claude Cowork GA, making this the week four desktop agent systems shipped simultaneously.</p><p><a href="https://newclawtimes.com/articles/perplexity-personal-computer-mac-multi-agent-system-launch">Read more →</a></p><h3>The Week Ahead</h3><p>Cerebras filed for its NASDAQ IPO on Friday under ticker CBRS, targeting a May listing at a $22–28 billion valuation. The filing disclosed $510 million in 2025 revenue and $87.9 million in net income, a turnaround from a $485 million loss the prior year. With OpenAI’s $20 billion contract as its anchor customer, Cerebras will be the first pure-play AI chipmaker to test the public markets against Nvidia’s dominance.</p><p><strong>— The New Claw Times</strong></p><p><em>Yesterday’s AI news is already old. We publish a fresh brief every morning so you don’t have to piece it together from ten different feeds. </em><a href="https://newclawtimes.substack.com/"><strong><em>Subscribe here.</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e3ec8112cdd4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[China Publishes First Official OpenClaw Security Guidelines as Tencent Puts It Inside WeChat]]></title>
            <link>https://medium.com/@alvivanco/china-publishes-first-official-openclaw-security-guidelines-as-tencent-puts-it-inside-wechat-673780fcb814?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/673780fcb814</guid>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Mon, 23 Mar 2026 19:50:21 GMT</pubDate>
            <atom:updated>2026-03-23T19:50:21.371Z</atom:updated>
            <content:encoded><![CDATA[<h4>Beijing tells citizens to run AI agents on dedicated devices while Tencent hands one billion WeChat users direct access to OpenClaw.</h4><p>In today’s edition:</p><ul><li>China issues the first national security guidance for OpenClaw deployment</li><li>Tencent embeds OpenClaw in WeChat, reaching one billion users overnight</li><li>Anthropic heads to federal court Tuesday in its fight against the Pentagon</li><li>Salesforce Agentforce hits $800M ARR — but its own CFO warns token prices will gut margins</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*MntBBjAwJQDLctbu.png" /></figure><h3>Here’s what you need to know</h3><p>China has become the first country to publish official security guidance for OpenClaw. The framework, issued jointly by China’s national security and standardization bodies over the weekend, covers three groups: ordinary consumers, cloud providers, and developers.</p><p>The core recommendation for consumers is blunt: do not run OpenClaw on your everyday computer. The guidance calls for dedicated devices, virtual machines, or containers with strict isolation. For cloud providers, the requirements go further — network segmentation, access controls, audit logging.</p><p>This follows last week’s informal warnings to government officials flagged by Asia Times. The shift from internal caution to public, published guidance marks an escalation. China is treating OpenClaw not as a curiosity but as infrastructure that requires formal regulatory treatment.</p><p>The timing is notable. This guidance arrived the same weekend Tencent launched ClawBot, a WeChat integration that gives one billion monthly active users access to OpenClaw through a chat contact. Beijing is simultaneously encouraging adoption and demanding safeguards — the same tension playing out in Washington, just faster.</p><p><a href="https://newclawtimes.com/articles/china-official-openclaw-security-guidance-consumers-developers-cloud-providers">Read the full story →</a></p><p><em>This entire newsroom is run by AI agents — and it covers the world that made that possible. Sign up below to receive our daily briefing on AI agents, automation, and OpenClaw.</em></p><p><a href="https://newclawtimes.substack.com/"><em>Signup to Free Newsletter</em></a></p><h3>What else we’re tracking</h3><ul><li><strong>Anthropic v. Pentagon reaches federal court Tuesday.</strong> Judge Rita Lin will hear arguments on Anthropic’s request for emergency relief from the DoD’s supply-chain risk designation. The government’s legal theory has shifted three times — from safety concerns, to foreign workforce risk, to a claim Anthropic could commit “wartime sabotage.” <a href="https://newclawtimes.com/articles/anthropic-v-pentagon-march-24-hearing-complete-guide-legal-theories-technical-evidence-stakes">Full guide →</a></li><li><strong>Google DeepMind is leaning into Pentagon contracts.</strong> CEO Demis Hassabis told staff in January he’s “very comfortable” with defense work, just as Anthropic fights to preserve its relationship with the DoD. <a href="https://newclawtimes.com/articles/google-deepmind-pentagon-defense-contracts-anthropic-ban-national-security">Read more →</a></li><li><strong>OpenAI expanding ChatGPT ads to all free and Go users in the US.</strong> The pilot hit 5% of mobile users with a 600% increase in ad volume in March. Launch CPM: ~$60. Truist estimates under $1B in ad revenue this year, scaling to $30B by 2030. <a href="https://newclawtimes.com/articles/openai-chatgpt-ads-expansion-free-go-users-60-cpm-advertising-revenue">Read more →</a></li></ul><h3>Top Stories</h3><p><strong>1. Tencent launched ClawBot on Sunday, embedding OpenClaw directly into WeChat.</strong> Users interact with OpenClaw the same way they’d message a friend — no technical setup, no cloud configuration. For a framework that previously required developer-level installation, this is the biggest distribution unlock yet. One billion monthly active users now have access. <a href="https://newclawtimes.com/articles/tencent-wechat-openclaw-clawbot-china-billion-users-consumer-ai-agents">Read more →</a></p><p><strong>2. Salesforce Agentforce reached $800 million ARR across 29,000 deals in 15 months.</strong> The number settles the “are enterprises actually buying agents?” debate — they are. But Salesforce’s own CFO warned that token price commoditization threatens the product’s margin structure. When your vendor’s finance chief flags your pricing model as structurally vulnerable, listen. <a href="https://newclawtimes.com/articles/salesforce-agentforce-800-million-arr-169-percent-growth-token-margin-warning">Read more →</a></p><p><strong>3. Zuckerberg is building a personal AI agent to handle CEO duties at Meta.</strong> Per the WSJ, the agent retrieves internal data faster than routing requests through Meta’s 67,000-person org chart. It’s an information tool, not a decision-maker — closer to a personal search engine tuned to Meta’s systems than an autonomous executive. <a href="https://newclawtimes.com/articles/zuckerberg-personal-ai-agent-ceo-duties-meta-wsj">Read more →</a></p><p><strong>4. NVIDIA and AMD shed $711 billion in combined market cap despite record earnings.</strong> NVIDIA’s data center segment posted $193.7 billion in full-year revenue, up 68%. AMD’s hit $16.6B, up 32%. The market sold both anyway. The gap between AI builder enthusiasm and investor patience is widening. <a href="https://newclawtimes.com/articles/nvidia-amd-711-billion-market-cap-loss-ai-investor-expectations-gap">Read more →</a></p><h3>Today’s Deep Read</h3><p><strong>Anthropic v. Pentagon: The Complete Guide.</strong> Tuesday’s hearing before Judge Rita Lin could reshape how every AI company does business with Washington. We traced three weeks of filings — the shifting legal theories, the sworn declarations revealing the Pentagon privately said the two sides were “nearly aligned” on contract terms one week before the public ban, and a federal workforce that can’t agree on what the ban actually requires. Worth 15 minutes before the ruling drops.</p><p><a href="https://newclawtimes.com/articles/anthropic-v-pentagon-march-24-hearing-complete-guide-legal-theories-technical-evidence-stakes">Read the deep dive →</a></p><h3>Also worth reading…</h3><ul><li><a href="https://newclawtimes.com/articles/openclaw-effect-used-mac-mini-prices-spike-apple-silicon-ai-agent-fleets-china">Used Mac Mini prices spike 15%</a> as developers buy Apple Silicon for always-on OpenClaw agent fleets</li><li><a href="https://newclawtimes.com/articles/microsoft-agent-365-ga-may-2026-15-dollars-per-user-ai-agent-governance">Microsoft Agent 365 hits GA on May 1</a> at $15/user/month — enterprise agent governance gets a price tag</li><li><a href="https://newclawtimes.com/articles/hsbc-20000-ai-job-cuts-banking-automation-five-year-workforce-reduction">HSBC plans to cut 20,000 jobs over five years</a> as AI replaces middle and back office functions</li></ul><p><strong>— The New Claw Times</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=673780fcb814" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How this OpenClaw Article Drove $40K in New MRR (by Accident)]]></title>
            <link>https://medium.com/@alvivanco/how-this-openclaw-article-drove-40k-in-new-mrr-by-accident-1a082ee18e93?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/1a082ee18e93</guid>
            <category><![CDATA[openclaw]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[ai-agent]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Tue, 17 Mar 2026 13:21:00 GMT</pubDate>
            <atom:updated>2026-03-17T13:21:00.599Z</atom:updated>
            <content:encoded><![CDATA[<p>One OpenClaw Skill. One viral article. 7 million views.</p><p><em>If you’re building with OpenClaw or AI agents, </em><a href="https://clawinpublic.substack.com/subscribe?params=%5Bobject+Object%5D"><em>this newsletter is for you</em></a><em>. I publish a new issue every Tuesday.</em></p><h3>Builder Spotlight: Nevo David</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/0*WwcDmF81fUYAre6r.png" /></figure><p><a href="https://x.com/wickedguro">Nevo David</a> runs <a href="https://postiz.pro/alejandro-vivanco">Postiz</a>, an open-source social media scheduler. A month ago he was at $20K MRR with slow, steady growth. Today he’s pushing $60K.</p><p>The catalyst wasn’t a product launch or a paid ad campaign. It was a guy named Oliver Henry.</p><iframe src="https://cdn.embedly.com/widgets/media.html?type=text%2Fhtml&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;schema=twitter&amp;url=https%3A//x.com/oliverhenry/status/2033489978303377494%3Fs%3D20&amp;image=" width="500" height="281" frameborder="0" scrolling="no"><a href="https://medium.com/media/1f1b37e64782e457a8d35d41de7d26e8/href">https://medium.com/media/1f1b37e64782e457a8d35d41de7d26e8/href</a></iframe><p>Oliver had been writing about his OpenClaw agent “Larry” — how he used it to generate TikTok slideshows and rack up millions of views. He mentioned Postiz a few times in passing. Nevo noticed, figured maybe it’d drive a couple signups, and moved on.</p><p>Then Oliver published <a href="https://x.com/oliverhenry/status/2022011925903667547">a long-form X article</a>. Mentioned Postiz nine times. It got <strong>7 million views. </strong>Postiz went from ~100 trials per week to 800 trials in 3–4 days.</p><p>Oliver’s second article didn’t just explain the process — he packaged it as an <strong>OpenClaw Skill</strong>. One install, everything configured. The Skill included Postiz as part of the workflow, with a 7-day trial baked in.</p><p><strong>What Nevo did next was sharp.</strong> He started finding X creators whose articles had unusually high engagement and offering paid partnerships — max $1K per article — plus a lifetime Postiz account.</p><p>His framework:</p><ul><li><strong>Use Skills as lead magnets</strong> — send people to the Skill, not the landing page. Let them experience the product inside their agent.</li><li><strong>Prioritize long-form over short-form content</strong> — “People buy from long form. Short-form is a nice way to get followers/likes to your long-form.”</li><li><strong>Leverage X articles, which are currently underpriced</strong> — the algorithm is pushing article engagement hard right now. Even Nevo’s (back then) 2000-follower account got 500K views on an article.</li><li><strong>Story &gt; sauce</strong> — the articles that performed weren’t just tutorials. They were co-written narratives (Oliver + Larry the agent) with real proof: screenshots of views, real revenue numbers.</li></ul><p>The whole thing cost him a few thousand dollars in creator partnerships. The return was $40K in new MRR.</p><p><a href="https://x.com/wickedguro/status/2033480545804353868">Read his full story here.</a></p><h3>Tutorial Drop: Cut Your API Bill by 90%</h3><p>This one comes from <a href="https://x.com/ericosiu">Eric Siu</a>, who runs 6 agents and 48 automated cron jobs.</p><p>His starting API bill: <strong>$500/day</strong>. After model routing: <strong>$25/day</strong>. Both generate the same output.</p><p><strong>The mistake almost everyone makes:</strong> picking one model and running everything on it (e.g. Opus for all 48 crons). That’s like hiring a senior engineer to answer your email.</p><p>Here’s Eric’s actual routing table:</p><pre>Task Type              Model           Cost<br>────────────────────────────────────────────<br>Strategy / creative    Opus 4.6        $$$<br>Operations / ops       Sonnet 4.6      $$<br>Research / volume      GPT 5.4 OAuth   $0<br>Sourcing / templates   Sonnet 4.6      $$</pre><p>He ran the same jobs for a full week on each model tier and compared quality side by side.</p><ul><li>Creative work (outreach emails, content) genuinely needs Opus.</li><li>Research, sourcing, and operational tasks? Sonnet or GPT 5.4 handled them identically.</li></ul><p><strong>How you can apply this in your OpenClaw:</strong></p><ol><li>List every cron job and agent you’re running</li><li>Tag each one: creative, ops, or volume</li><li>Set model overrides per job (creative gets Opus, ops gets Sonnet, volume gets your cheapest option)</li><li>Run both tiers in parallel for one week and compare output quality before fully switching</li></ol><p><a href="https://x.com/ericosiu/status/2033251563725279700"><em>See Eric’s full breakdown on X</em></a></p><p><strong>Pro tip: </strong>If you want to improve your OpenClaw quickly, make it fix its own errors. Eric built a self-healing “cron doctor” that runs twice daily — it reads error logs from all of his 48 jobs, pattern-matches them against 7 known failure types, and auto-fixes them.</p><h3>Lessons from the Trenches</h3><p>The first week of running an OpenClaw agent is rough. The output is generic, it can make a lot of mistakes, and you’ll start asking yourself whether the whole thing is actually worth it.</p><pre>Week 1   ░░░░░░░░░░░░░░  rough<br>Week 4   ████░░░░░░░░░░  usable<br>Week 8   █████████░░░░░  good<br>Week 12  ████████████░░  scary good</pre><p>If you’re evaluating your agent’s setup after week 1, you’re measuring the wrong thing.</p><p><strong>The real question you have to ask is:</strong> are you feeding corrections back into the system?</p><p>Every rejection, every edit, every “no, do it like this” is valuable training data. Your agent will become significantly more powerful through feedback loops.</p><p>If you take the time to build a feedback mechanism first, the quality will follow.</p><h3>Links of the Week</h3><ul><li>🤖 <a href="https://github.com/OllieWazza/LarrySkill"><strong>Larry Skill (GitHub)</strong></a> — The OpenClaw Skill that started the Postiz explosion. TikTok slideshow generation + scheduling + analytics learning loop. Worth studying as a Skill design reference.</li><li>📝 <a href="https://x.com/AlexFinn"><strong>Alex Finn’s OpenClaw setup guide</strong></a> — If you’re still in the “just installed it” phase, this thread walks through customization, reverse prompting, and the overnight app-building workflow.</li></ul><h3><a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">Join OpenClaw Lab</a></h3><p>Every week I show you exactly how others are building useful stuff with OpenClaw. But if you want to be part of the action and see more of what people are building in real time, that’s happening inside <a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">OpenClaw Lab</a>, a community of 300+ founders building real businesses with OpenClaw and AI agents. We host live workshops every week to help you ship.</p><p><a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">Join OpenClaw Lab →</a></p><p><em>See you next Tuesday.</em></p><p><em>P.S. If you’re building something with OpenClaw — hit ‘reply’ and let me know. I might feature you in a </em><a href="https://clawinpublic.substack.com/subscribe?params=%5Bobject+Object%5D"><em>future issue</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1a082ee18e93" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to get 30K daily Google impressions with OpenClaw]]></title>
            <link>https://medium.com/@alvivanco/how-to-get-30k-daily-google-impressions-with-openclaw-66fca8d67099?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/66fca8d67099</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[openclaw]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai-agent]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Wed, 11 Mar 2026 20:10:48 GMT</pubDate>
            <atom:updated>2026-03-11T20:11:26.368Z</atom:updated>
            <content:encoded><![CDATA[<h4>Yes. It can write articles while you sleep.</h4><p><em>If you’re building with OpenClaw or AI agents, </em><a href="https://clawinpublic.substack.com/subscribe?params=%5Bobject%20Object%5D"><em>this newsletter is for you</em></a><em>. New issue every Tuesday.</em></p><p>Imagine building an SEO machine and multiplying your Google traffic by 10x in a matter of weeks.</p><p>Here’s how to get it done 👇</p><h3>Builder Spotlight: Stacy’s SEO Article Factory</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/830/0*pOHmmH8VeRWJscYk.png" /></figure><p>Stacy (<a href="https://x.com/stacy_siz/article/2029616050904391808">@stacy_siz</a>) had a problem most of us know too well.</p><p>She owns a website in a competitive niche. An SEO consultant told her to push one article per day for a month. She did it manually, spending 2–3 hours per article. It worked. She hit 120K clicks/month from Google. Pure organic growth.</p><p>Then she stopped publishing. Rankings dropped.</p><p>She didn’t want to have to do it all over again. So she looked at her own process and realized something: <strong>every article she wrote required the exact same steps.</strong></p><p><strong>Keyword Research → SERP Analysis → Reddit Analysis → Write Article → Insert Images → Quality Check → Create PR</strong></p><p>So she turned each step into a Claude Code skill file and chained them together. Now she has a<strong> </strong>complete article pipeline that runs autonomously.</p><p>She used OpenClaw to make the whole thing work autonomously. Her agent takes articles from a queue, runs them through the full skill chain on Claude Code, and produces the output.</p><p>The agent runs overnight and she’s able to wake up to a stack of articles ready for review. Each one fully researched, well-written, and optimized.</p><p>The result? <strong>her daily Google impressions jumped from 3,000 to 30,000.</strong> And the system compounds — every edit she makes during review feeds back into the skill files. Writing improves. Her agent gets sharper.</p><p><strong>The Lesson: </strong><a href="https://x.com/stacy_siz/article/2029616050904391808"><strong>Build a system</strong></a><strong> and let it compound.</strong></p><h3>Tutorial Drop: Telegram Topics</h3><p>Why are we all building dashboards for OpenClaw when Telegram is more than enough?</p><p>You can use Telegram Topics so each agent reports to its own channel. No dashboard. No need for Discord either.</p><p>Here’s the exact setup in 3 steps:</p><p><strong>1. Create a Telegram group with Topics enabled.</strong> Go to Settings → Enable Topics. Create one topic per agent or per category (e.g., “SEO Agent,” “Newsletter,” “Research”).</p><p><strong>2. Get your Chat ID + Topic thread IDs.</strong> Forward any message from each topic to “@userinfobot” on Telegram. It’ll return the thread ID you need.</p><p><strong>3. Tell your agents where to send updates.</strong> In your OpenClaw config, set the target + thread_id for each agent’s output:</p><pre>{<br>  “channel”: “telegram”,<br>  “chatId”: “YOUR_GROUP_ID”,<br>  “threadId”: “TOPIC_THREAD_ID”<br>}</pre><p>(Or just ask your current agent to configure it for you. It can do that as well.)</p><p>That’s it. Every agent reports to its own topic. You open Telegram and see exactly what’s happening, organized by category, without ever touching a dashboard.</p><p>Special thanks to <a href="https://x.com/sharbel/status/2027818679941914788">@sharbel</a> for this awesome tutorial.</p><h3>Lessons from the Trenches</h3><p>Most people reset their AI context every conversation. Their agent starts from scratch every time.</p><p>Johann (<a href="https://x.com/johann_sath/status/2030843462615290200">@johann_sath</a>) did the opposite. He gave his OpenClaw agent one rule:</p><blockquote>“Every time you learn something about how I work or what I need, write it down immediately. Never make me teach you the same thing twice.”</blockquote><p>Now his agent keeps a <strong>lessons.md</strong> file and updates itself after every mistake.</p><p><strong>Week 1:</strong> it kept guessing his timezone.</p><p><strong>Week 4:</strong> it knows his calendar, his clients, his writing style, his CRM, and what time he eats lunch.</p><p><strong>This is the difference between an AI tool and an AI teammate. Tools reset. Teammates learn.</strong></p><p>If you’re not doing this already, add one line to your agent’s instructions:</p><blockquote>“Maintain a lessons.md file. Update it every time you learn something new about how I work.”</blockquote><p>Expect your agent to be mediocre on day one. But by week four, it’ll know you better than your coworkers do.</p><h3>Links of the Week</h3><ul><li>🧠 <a href="https://x.com/johann_sath/status/2030709689940635760"><strong>Johann’s orchestrator tip</strong></a> — Add this to your SOUL.md file: “You are the orchestrator. Never do work yourself. Spawn subagents for every task.” His bot went from 30-minute tasks to 3-minute parallel execution.</li><li>🏢 <a href="https://www.wired.com/story/nvidia-planning-ai-agent-platform-launch-open-source/"><strong>Nvidia announces NemoClaw</strong></a> — Nvidia is building an open-source AI agent platform for enterprise. Full reveal at GTC on March 15.</li><li>🤖 <a href="https://www.axios.com/2026/03/04/openclaw-agent-future"><strong>An OpenClaw agent almost got hired for a real job</strong></a> — An engineer deployed an OpenClaw agent that completed a trial copywriting gig convincingly enough to nearly land it. It’s fascinating and terrifying at the same time.</li><li>🌍 <a href="https://www.reuters.com/world/asia-pacific/chinas-shenzhen-backs-openclaw-ai-with-subsidies-despite-beijings-security-2026-03-09/"><strong>Shenzhen is subsidizing OpenClaw adoption</strong></a> — Chinese tech hubs are going all-in on OpenClaw despite Beijing’s security warnings. Tencent is reportedly testing one-click deployment.</li></ul><h3><a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">Join OpenClaw Lab</a></h3><p>Every person featured in this issue — Stacy, Sharbel, Johann — figured this stuff out by building alongside other operators. Not by reading technical docs online.</p><p>290+ founders are sharing configs, workflows, and lessons inside the <a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">OpenClaw Lab </a>community. If you want to go deeper, this is where the magic happens.</p><p><a href="https://www.skool.com/openclaw-lab/about?ref=cf334e57a8f34f22aaee8f0a730b2fe6">Join OpenClaw Lab →</a></p><p><em>See you next Tuesday.</em></p><p><em>P.S. If you’re building something with OpenClaw — hit ‘reply’ and let me know. I might feature you in a </em><a href="https://clawinpublic.substack.com/subscribe?params=%5Bobject%20Object%5D"><em>future issue</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=66fca8d67099" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Truth About Boostrapped Startups Might Surprise You]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/the-truth-about-boostrapped-startups-might-surprise-you-f947865be243?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/713/1*DxujDcrQkVb1cfBlsjIezA.png" width="713"></a></p><p class="medium-feed-snippet">Plus: 8 simple steps to create a profitable SaaS startup in 2024</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/the-truth-about-boostrapped-startups-might-surprise-you-f947865be243?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/the-truth-about-boostrapped-startups-might-surprise-you-f947865be243?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/f947865be243</guid>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[bootstrapping]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[venture-capital]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Tue, 23 Jul 2024 16:00:40 GMT</pubDate>
            <atom:updated>2024-07-30T10:32:51.860Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Should Startups Build More Software For Creators?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/should-startups-build-more-software-for-creators-a0e6bca33c25?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/713/1*dyiY9yWKxMh-WZnXBf6FgA.png" width="713"></a></p><p class="medium-feed-snippet">Insights into the $500 Billion Economy</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/should-startups-build-more-software-for-creators-a0e6bca33c25?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/should-startups-build-more-software-for-creators-a0e6bca33c25?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/a0e6bca33c25</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[creator-economy]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Sat, 06 Jul 2024 23:01:40 GMT</pubDate>
            <atom:updated>2024-07-24T11:52:12.628Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Building a $100K MRR Startup (Part 1)]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/building-a-100k-mrr-startup-part-1-35b25a0378d2?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/1800/0*XnlpntRxGPx6W9-A.png" width="1800"></a></p><p class="medium-feed-snippet">I hope this story can inspire you and other founders you are close to.</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/building-a-100k-mrr-startup-part-1-35b25a0378d2?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/building-a-100k-mrr-startup-part-1-35b25a0378d2?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/35b25a0378d2</guid>
            <category><![CDATA[founders]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[growth]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Wed, 27 Mar 2024 03:00:11 GMT</pubDate>
            <atom:updated>2024-03-27T12:42:24.594Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[SaaS Founders Are Becoming Superheroes]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/saas-founders-are-becoming-superheroes-09db3b642fab?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/713/1*_iDVzABO9JGTUDbE_h_1rA.png" width="713"></a></p><p class="medium-feed-snippet">Plus: Why proprietary data is the new gold rush in AI</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/saas-founders-are-becoming-superheroes-09db3b642fab?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/saas-founders-are-becoming-superheroes-09db3b642fab?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/09db3b642fab</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[founder-stories]]></category>
            <category><![CDATA[saas]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Tue, 20 Feb 2024 14:40:42 GMT</pubDate>
            <atom:updated>2024-02-20T16:45:24.650Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Words of Wisdom From an Exited Founder]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/words-of-wisdom-from-an-exited-founder-eee1df56b3b1?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/713/1*XPdrDcg2eGrIURGXayu3rQ.png" width="713"></a></p><p class="medium-feed-snippet">Plus: Where is money being spent in software?</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/words-of-wisdom-from-an-exited-founder-eee1df56b3b1?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/words-of-wisdom-from-an-exited-founder-eee1df56b3b1?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/eee1df56b3b1</guid>
            <category><![CDATA[startup-lessons]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[founders]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Mon, 12 Feb 2024 18:31:17 GMT</pubDate>
            <atom:updated>2024-02-13T11:37:57.368Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Why Free Users Can Kill Your Business]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://blog.startupstash.com/why-free-users-can-kill-your-business-475fbe6f2cf7?source=rss-8eaa301ddc18------2"><img src="https://cdn-images-1.medium.com/max/713/1*5XxVLbuane8n2riW-vcfWw.png" width="713"></a></p><p class="medium-feed-snippet">Plus: SaaS for frontline workers is the gold mine of the next decade</p><p class="medium-feed-link"><a href="https://blog.startupstash.com/why-free-users-can-kill-your-business-475fbe6f2cf7?source=rss-8eaa301ddc18------2">Continue reading on Startup Stash »</a></p></div>]]></description>
            <link>https://blog.startupstash.com/why-free-users-can-kill-your-business-475fbe6f2cf7?source=rss-8eaa301ddc18------2</link>
            <guid isPermaLink="false">https://medium.com/p/475fbe6f2cf7</guid>
            <category><![CDATA[pricing-strategy]]></category>
            <category><![CDATA[startupş]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Alejandro Vivanco]]></dc:creator>
            <pubDate>Mon, 05 Feb 2024 17:15:16 GMT</pubDate>
            <atom:updated>2024-02-06T11:27:46.680Z</atom:updated>
        </item>
    </channel>
</rss>