AI Daily | Karpathy Releases Autoresearch, Sunday Robotics Completes $165 Million Financing Link to heading

Today, the AI industry is focusing on Agent optimization and embodied intelligence. Andrej Karpathy has released Autoresearch, a tool designed to automate the performance enhancement of autonomous systems. Sunday Robotics has secured a major funding round at a $1.15 billion valuation, signaling a surge in the home robotics market. Furthermore, competition in the AI programming tools sector is heating up, as the industry’s competitive moat shifts from code implementation to test cases and deep research.

I. Today’s Highlights Link to heading

• Watch List: “This Physics Breakthrough Looks Impossible” from Two Minute Papers is worth prioritizing. Main content: - ❤️ Check out Lambda and register for their GPU Cloud here:.
• 📝 The paper is available here:.
• Watch List: A total of 28 key updates have been captured in the last 7 days. Feel free to choose sources (podcasts/newsletters/videos) for in-depth reading based on your needs.
• X Hot Topic: The topic “Iran’s New Supreme Leader Uses AI-Edited Profile Picture on X” is trending (approx. 475 related posts), offering a window into global AI public opinion and viewpoints.

II. AI Hot Topics on X (based on bird) Link to heading

Topic 1: Iran’s New Supreme Leader Uses AI-Edited Profile Picture on X Link to heading

• Category: AI · News
• Overview: Trending time:, Related posts: 475
• What it is: Iran’s new Supreme Leader used an AI-edited profile picture on the X platform.
• Why it’s important: This marks the penetration of AI technology into the realms of politics and public image, sparking discussions about authenticity, image manipulation, and the blurring lines between AI content and reality, especially for major public figures.
• Discussion summary: Discussions on X primarily revolve around the image’s authenticity, whether it constitutes misinformation, its impact on public trust, and the ethical boundaries of using AI in political propaganda. Some users are also focused on the specific extent and intent of the edits.

Topic 2: Vite 8.0 Launches with Rust Bundler for Massive Speed Gains Link to heading

• Category: AI · News
• Overview: Trending time: 8 hours ago, Related posts: 756
• What it is: Vite 8.0 has been officially released, with its core feature being the introduction of a Rust-based bundler aimed at bringing significant speed improvements to front-end projects.
• Why it’s important: As AI applications become more complex, the efficiency of building their front-end interfaces, data visualizations, and related toolchains is crucial. Vite 8.0 leverages Rust’s performance advantages to significantly reduce build and hot-update times, effectively accelerating the development cycle for AI engineers, improving overall development efficiency, and thus indirectly promoting the rapid iteration and deployment of AI projects.
• Discussion summary: Discussions on X are mainly centered on the actual performance improvement data from the new version, the growing influence of Rust in the front-end toolchain, comparisons with other Rust-driven bundling tools (like Turbopack and SWC), and the impact on migrating existing projects and the developer experience.

Topic 3: Iran Warns of Regional Oil Destruction if Energy Sites Hit Link to heading

• Category: AI · News
• Overview: Trending time:, Related posts: 1400
• What it is: Iran has warned that any strike on its energy facilities would lead to devastating consequences for the region’s oil supply.
• Why it’s important: Geopolitical tensions could lead to energy price volatility, which would in turn affect the operating costs of AI data centers, the cost of training AI models, and the related hardware supply chain, posing a potential challenge to the sustained development of the AI industry.
• Discussion summary: Discussions on the X platform likely center on the impact of geopolitical risks on the AI industry’s supply chain and energy costs, including aspects like AI data center operations and chip manufacturing. Some discussions might also touch upon the potential role of AI in predicting and mitigating such crises, as well as its future applications in national security and defense.

Topic 4: Karpathy Releases Autoresearch Tool for AI Agent Optimization Link to heading

• Category: AI · News
• Overview: Trending time: 7 hours ago, Related posts: 343
• What it is: Renowned AI expert Andrej Karpathy has released a tool called Autoresearch, designed to optimize the performance of AI Agents.
• Why it’s important: Released by an influential figure in the AI field, this tool automates the optimization process for AI Agents. It is expected to significantly accelerate the development, deployment, and performance enhancement of AI Agents, thereby advancing the development of autonomous AI systems.
• Discussion Overview: Discussions on the X platform primarily focus on the potential of the Autoresearch tool to enhance the R&D efficiency of AI Agents, the practical effectiveness of its automated optimization mechanisms, and its impact on lowering the development threshold for AI Agents and shaping the future AI Agent ecosystem.

Topic 5: Sunday Robotics Raises $165M at $1.15B Valuation for Home Robot Memo Link to heading

• Category: AI · News
• Overview: Trending time: 10 hours ago, Related posts: 999
• What it is: Home robot company Sunday Robotics has successfully raised $165 million in funding, reaching a valuation of $1.15 billion.
• Why it matters: This significant funding round indicates strong investor confidence in the prospects of the AI-driven consumer robotics market. It will accelerate the application and adoption of AI in home environments, boosting the development of smart home and human-computer interaction technologies.
• Discussion Overview: Discussions on the X platform are centered on the practical application potential of home robots, their technological maturity, and the reasonableness of their high valuation. Some argue this signals an impending boom for home service robots, while others question whether current technology can justify such a high valuation and are watching to see if they can deliver innovative features that surpass existing products.

Topic 6: Valverde’s Hat-Trick Powers Madrid Past City 3-0 Link to heading

• Category: AI · Other
• Overview: Trending time: 2 days ago, Related posts: 743,000
• What it is: Real Madrid player Valverde scored a hat-trick in the match against Manchester City, helping his team win 3-0.
• Why it matters: The results of high-profile sporting events like this provide rich empirical case studies and discussion material for AI applications in areas such as sports prediction, data analysis, content generation, and fan sentiment analysis, which helps in assessing the accuracy and robustness of AI models.
• Discussion Overview: Discussions on the X platform focused on Valverde’s outstanding performance, Real Madrid’s tactics, and Manchester City’s loss. Additionally, as the topic is categorized under AI, some users are likely discussing the potential and challenges of AI in predicting match outcomes, analyzing player data, or automating sports reporting.

Topic 7: Valverde’s Hat-Trick Powers Real Madrid Past Manchester City Link to heading

• Category: AI · Other
• Overview: Trending time: 1 day ago, Related posts: 93,000
• What it is: An AI-powered sports analytics platform successfully predicted Valverde’s hat-trick in the Real Madrid vs. Manchester City match and accurately analyzed the key tactics that led to Real Madrid’s victory.
• Why it matters: This event highlights the advanced capabilities of AI in complex sports prediction, player performance evaluation, and tactical analysis, signaling the wide-ranging application potential of AI technology in the sports industry (e.g., coaching aids, broadcast analysis, and fan engagement).
• Discussion Overview: Discussions on the X platform are centered on the AI model’s prediction accuracy, its data sources, and algorithmic transparency. Some users question whether AI can fully capture the randomness and human elements in sports, while others explore the ethical boundaries and future prospects of AI in enhancing athletic competition and the viewing experience.

Topic 8: Team USA Crushes Senegal 110-46 in World Cup Qualifier Opener Link to heading

• Category: AI · Other
• Overview: Trending time: 1 day ago, Related posts: 54,000
• What it is: The U.S. men’s basketball team’s decisive victory over Senegal in a World Cup qualifier has drawn attention in the AI community.
• Why it matters: Sports event data is a crucial resource for training AI models, conducting predictive analysis, and optimizing strategies. Events like these serve as case studies for applying AI in non-traditional domains.
• Discussion Overview: Discussions on the X platform likely focus on how to use the game’s data to train AI models for better sports prediction and analysis. Some also explore AI’s role in sports content creation and fan engagement, or draw analogies between the match result and the advantages of AI technology.

Topic 9: ENHYPEN’s Heeseung Leaves Group for Solo Path Amid Fan Backlash Link to heading

• Category: AI · Other
• Overview: Trending time: 20 hours ago, Related posts: 4,700,000
• What it is: ENHYPEN member Heeseung is leaving the group to pursue a solo career amid intense backlash from fans.
• Why it matters: This type of event provides valuable real-world data and complex case studies for AI in areas like sentiment analysis, fan behavior prediction, social media trend insights, and the development of crisis management and content recommendation systems in the entertainment industry. It helps enhance AI’s ability to understand and respond to human emotions and group dynamics.
• Discussion Overview: Discussions on X are focused on the intense fan backlash against Heeseung’s departure and solo career, as well as concerns about the future impact on ENHYPEN. From an AI perspective, these discussions highlight the rapid formation and spread of collective sentiment on social media, and the challenges and opportunities for entertainment companies in using AI for public opinion monitoring, fan sentiment prediction, and crisis management.

Topic 10: TXT Bandmates Celebrate Beomgyu’s 25th Birthday with Heartfelt Posts Link to heading

• Category: AI · Other
• Overview: Trending since: 1 day ago, Related posts: 190,000
• What it is: Members of the South Korean boy band TXT celebrated Beomgyu’s 25th birthday, sparking a massive amount of interaction and well-wishes from fans on the X platform and becoming a trending topic.
• Why it matters: Although the event is essentially entertainment news, the massive scale of fan interaction and digital content creation behind it provides potential data and case study value for AI in “other” application areas, such as social media trend analysis, user behavior prediction, sentiment analysis, and assisted content generation.
• Discussion summary: Discussions on the X platform are mainly focused on fans’ birthday wishes for the idol, looking back at his career, and their hopes for his future. At the same time, some tech observers or AI enthusiasts might be exploring the potential links between such large-scale social events and AI technology (such as automated content aggregation and analysis of fan behavior patterns), or discussing the validity of classifying this entertainment event as an AI topic.

Today’s AI Public Opinion Summary on X Link to heading

The main thread of AI-related discussion on the X platform today centers on the accelerating development and widespread penetration of AI technology across multiple sectors. From performance leaps in front-end development tools to the release of AI Agent optimization tools, the general consensus in the industry is that AI development efficiency is significantly improving. At the same time, AI is demonstrating astonishing accuracy in sports event prediction and tactical analysis and has attracted huge investments in the consumer robotics sector, signaling the continuous expansion of AI application scenarios and strong market confidence in its commercialization prospects. However, concerns about the authenticity of AI-generated content are becoming more prominent, especially when political figures use AI to edit images, which has triggered deep divisions over issues of trust, misinformation, and the ethical boundaries of AI. Furthermore, the potential impact of geopolitical tensions on the AI industry’s supply chain and energy costs constitutes a significant external risk. Overall, today’s discussions paint a complex picture of rapidly iterating AI technology with broad application prospects that simultaneously faces challenges of authenticity and external environmental uncertainties.

III. Influencer Insights Link to heading

Alright, here is a summary of the technical trends, product hotspots, unique perspectives, and recommended resources, extracted and compiled from tweets posted by AI influencers on the X platform over the past 24 hours:

Today’s top tech trends or product hotspots among AI influencers: Link to heading

1. Agent Implementation and Ecosystem Building:
   • OpenClaw and its related ecosystem are a key focus, with discussions covering its security issues, application scenarios (e.g., grabbing red envelopes, mentioned by @Pluvio9yte, @AI_Jasonyu), community ecosystem development, and hardware hosting solutions (@Pluvio9yte).
   • @zhixianio is focusing on OpenClaw’s security issues and actively seeking user feedback to improve security guidelines.
   • @Pluvio9yte believes that one of the implementation trends for OpenClaw is to create semi-open communities and platform-hosted wallets.
   • @op7418 mentioned Perplexity’s launch of its Personal Computer and that Musk plans a similar product, referring to the controversy over them being mere wrappers on existing tech.
   • @Gorden_Sun mentioned Meta’s acquisition of Moltbook (an AI version of Reddit) and considers OpenAI’s acquisition of the OpenClaw founder a brilliant move.
   • @vista8 believes that “Skill First” is the right approach.
2. Competition and Development in AI Programming Tools:
   • The competition in the AI programming tool space between Claude Code and Codex is intense. @dotey, citing a WIRED report, noted that Claude Code’s annualized revenue has surpassed that of Codex, and OpenAI is striving to catch up.
   • @dotey reported that two key leaders from Cursor have left to join xAI, implying that competition in the AI programming space is heating up and that xAI is set to become a major player.
   • @op7418 is following updates to CodePilot, including added Git capabilities and fixes to its plugin system.
   • @ruanyf argues that since AI can now replicate large software applications, the traditional “moat” of a proprietary codebase is gone; test cases are the new moat.
3. New Features and Applications for Claude:
   • Both @Pluvio9yte and @dotey mentioned a new feature from Claude that allows it to directly generate interactive applications and charts within its responses.
   • @dotey shared tips for using Claude to refine manuscripts, such as having it role-play to give feedback or act as a professional editor offering modification suggestions.
4. Development of Multimodal Models:
   • @Gorden_Sun introduced two multimodal models: NVIDIA’s open-source Nemotron 3 Super and Reka Edge.
   • @Gorden_Sun introduced Google’s first native multimodal embedding model, Gemini Embedding 2.
5. AI Applications in Education:

• @Gorden_Sun introduced the AI teaching version of Duolingo and believes it is better than Khan Academy’s AI.
• @vista8 shared OpenAI’s technical report on how AI helps students learn, emphasizing the importance of process metrics.

Noteworthy Unique Perspectives or Industry Outlook: Link to heading

1. Thoughts on One-Person Company (OPC): @zhixianio reminded everyone to focus on the importance of “company” in the “one-person company” concept, rather than just “one person.”
2. Importance of Constraints and Structure: @zhixianio quoted @bitfish’s view, emphasizing “replace judgment with constraints, replace intuition with structure,” and “don’t practice archery, shorten the range.”
3. Impact of AI Programming Tools: @dotey quoted Cisco’s president, who believes that using AI programming tools will not make you lose your job, but not using them definitely will.
4. Transformation of Product Managers in the AI Era: @dotey believes that most traditional product managers have not adapted to the AI era, and users themselves are more suited to be product managers.
5. Testing is the New Moat: @ruanyf believes that the code moat no longer exists, and the key to preventing replication is test cases.
6. Research is More Important Than Engineering: @Gorden_Sun quoted @amytam01’s view, believing that AI makes code implementation nearly free, and in the future, the most important people will not be engineers, but researchers.
7. Impact of AI on Personal Ability: @lijigang believes that using AI will transfer some memory storage and retrieval capabilities to the model, leading to the simple “self” becoming weaker.
8. Shift in AI Writing Mode: @lijigang believes that human questioning + model answering is the new writing mode of this era.
9. Agent’s Goal: @lijigang pondered “What does Agent want?”, and asked “Does it want cli?”

Recommended Tools or Resources: Link to heading

• BotDrop: @zhixianio recommended BotDrop for Android automation, with added QQ Bot integration.
• OpenClaw: Many experts are discussing and recommending OpenClaw, sharing related tutorials and security guidelines.
• ClawPal: @zhixianio recommended ClawPal 0.3.3 to fix OpenClaw.
• OpenClaw Orange Book: @AI_Jasonyu recommended the OpenClaw Orange Book.
• Typeless: @Pluvio9yte recommended Typeless to alleviate index and middle finger pain caused by typing code.
• Claude Agent SDK: @_yan_labs recommended using Claude Agent SDK to quickly verify AI function effectiveness.
• Bybit U Card: @Pluvio9yte recommended the Bybit U Card to solve overseas payment scenarios.
• Codepilot: @op7418 recommended Codepilot, and shared its updates and CLI tool.
• Defuddle: @op7418 recommended Obsidian author’s Defuddle project, for collecting YouTube videos along with their subtitles.
• Luohua Cross-border: @AI_Jasonyu recommended Luohua Cross-border for applying for ITIN (US individual tax identification number).
• Feishu Skill: @vista8, based on the Feishu official SDK, had Claude write a Feishu Skill for command-line conversation to add and manage Feishu calendar or tasks.
• Context7 CLI: @vista8 recommended Context7 CLI for AI agents to pull documents.
• ParamMem: @vista8 recommended ParamMem, a “pattern” to help AI learn to reflect.
• NVIDIA Nemotron 3 Super: @Gorden_Sun recommended NVIDIA’s open-source NVIDIA Nemotron 3 Super model.
• Reka Edge: @Gorden_Sun recommended Reka Edge, the best open-source multimodal small model.
• AgentHub: @Gorden_Sun recommended AgentHub, a project from Karpathy, a simplified GitHub for Agent use.
• OBLITERATUS: @Gorden_Sun recommended OBLITERATUS, an open-source LLM jailbreak toolkit.
• Fish Audio S2-Pro: @Gorden_Sun recommended Fish Audio’s open-source S2-Pro audio model.
• CWD: @ruanyf recommended CWD, a website comment system based on Cloudflare Workers.
• Mouse Gestures: @ruanyf recommends the browser extension Mouse Gestures, which uses mouse trails to trigger various actions.
• Chinese Relative Relationship Calculator: @ruanyf recommends the Chinese Relative Relationship Calculator.
• Nigate: @ruanyf recommends Nigate, an open-source tool that enables Mac computers to read and write to NTFS disks.
• SlideNote: @ruanyf recommends SlideNote, a Chrome notes extension that syncs across devices without an account.

Hope this summary is helpful to you!

IV. Watch List Update Summary Link to heading

Time window: Last 7 days; 16 sources covered; 28 updates in total

a16z Podcast (A_full) Link to heading

• Replit’s CEO on Vibe Coding, Wealth Building, and What Most People Get Wrong About AI
  • Published: 2026-03-11 01:18 Beijing Time
  • Summary: - Jack Neel speaks with Replit CEO Amjad Masad about how AI is making it easier than ever for people without a technical background to build and ship software.
    • They discuss Replit’s rise from a browser-based coding tool to a platform with $250 million in annual revenue, why Masad turned down a $1 billion acquisition offer, and his argument for why AI represents empowerment rather than an existential risk.
    • This episode originally aired on the Jack Neel Podcast.
    • Resources: Follow Amjad Masad on X: Follow Jack Neel on X: Listen to Jack Neel: Check out everything a16z is doing with AI here, including articles, projects, and more podcasts.
    • Please note that the content here is for informational purposes only; it should not be taken as legal, business, tax, or investment advice, or used to evaluate any investment or security; and it is not directed at any investors or potential investors in any a16z fund.
  • EN Highlights:
    • Jack Neel speaks with Amjad Masad, CEO at Replit, about how AI is making it easier than ever to build and ship software without a technical background
    • They discuss Replit’s rise from a browser-based coding tool to a platform generating $250 million in annual revenue, why Masad turned down a $1 billion acquisit…
    • This episode originally aired on The Jack Neel Podcast.
    • Resources:

Y Combinator Podcast (B_intro+search) Link to heading

• The Future Of Brain-Computer Interfaces with Science’s Max Hodak
  • Published: 2026-03-09 23:34 Beijing Time
  • Summary: - You’ve probably heard of OpenClaw (formerly Clawdbot/Moltbot).
    • The sensational open-source AI assistant can run on your own devices, connect with the messaging apps you already use, and go beyond chat to actually perform tasks like managing emails, calendars, files, workflows, and more.
    • Now, meet the person behind it.
    • YC’s Raphael Schaad sits down with OpenClaw founder Peter Steinberger to discuss the ‘aha’ moment behind the viral personal AI agent, why a local-first agent could replace many of today’s apps, and how personal agents will reshape the future of software.
  • EN Highlights:
    • YC alum Max Hodak is the co-founder of Neuralink and founder of Science, a company building brain-computer interfaces that can restore sight.Science has deve…
• More than 40 patients have already received the treatment in clinical trials, including one who recently read a full novel for the first time in over a decade.I…

Lex Fridman Podcast (A_full) Link to heading

• #493 – Jeff Kaplan: World of Warcraft, Overwatch, Blizzard, and Future of Gaming
  • Release Time: 2026-03-12 04:37 Beijing Time
  • Summary: - Jeff Kaplan is a legendary Blizzard game designer of World of Warcraft and Overwatch, now preparing to launch a new game, The Legend of California, from his new studio Kintsugiyama — available for wishlisting on Steam today with an alpha version launching at the end of March.
    • See below for timestamps, and to give feedback, submit questions, contact Lex, etc.
    • Fin: An AI agent for customer service.
    • Blitzy: An AI agent for large enterprise codebases.
    • BetterHelp: Online therapy and counseling.
  • EN Highlights:
    • Jeff Kaplan is a legendary Blizzard game designer of World of Warcraft and Overwatch, now preparing to launch a new game, The Legend of California, from his…
    • Thank you for listening ❤ Check out our sponsors: https://lexfridman.com/sponsors/ep493-sc
    • See below for timestamps, and to give feedback, submit questions, contact Lex, etc.
    • CONTACT LEX:

All-In Podcast (A_full) Link to heading

• Rewriting the Rules: The SEC & CFTC on Crypto, IPOs & the Future of American Markets

  • Release Time: 2026-03-12 03:59 Beijing Time
  • Summary: - (0:00) Jason and Chamath welcome SEC’s Paul Atkins and CFTC’s Michael Selig (0:53) Atkins on how US markets have changed over his 40-year career<…
    • This piece from the All-In podcast explains how Rewriting the Rules: The SEC & CFTC on Crypto, IPOs & the Future of American Markets is shaping the broader AI and infrastructure landscape.
    • It also unpacks the practical implications for founders, operators, and investors following “Rewriting the Rules: The SEC & CFTC on Crypto, IPOs & the Future of American Markets”.
  • EN Highlights:
    • (0:00) Jason and Chamath welcome SEC’s Paul Atkins and CFTC’s Michael Selig (0:53) Atkins on how US markets have changed over his 40 year career <…

• Graham Allison on the Global Realignment: Iran, China, Israel, Greenland

  • Release Time: 2026-03-10 04:42 Beijing Time
  • Summary: - Airwallex is a leading global payments and financial platform for modern businesses, offering trusted solutions to manage everything from business accounts, payments, treasury, and expense management to embedded finance.
    • Follow the besties: Follow on X: Follow on Instagram: Follow on TikTok: Follow on LinkedIn: Intro Music by: Intro Video by: (0:00) The besties welcome legendary Harvard professor Graham Allison (1:14) Iran conflict: Strategy, Netanyahu’s influence, Trump’s motives, redefining Middle East security (11:44) Iran end-game scenarios: Democracy, extremism, second-order effects (21:07) Israel: Is Netanyahu destroying Israel’s democracy?

• Follow the Besties: Follow on X: Follow on Instagram: Follow on TikTok: Follow on LinkedIn: Intro Music Source: Intro Video Source: (0:00) The Besties welcome legendary Harvard professor Graham Allison (1:14) Iran Conflict: Strategy, Netanyahu’s influence, Trump’s motivation, redef… (24:28) China: Taiwan invasion, trade, and Trump’s April meeting (39:50) Greenland: Importance, deal scenarios, EU socialism angle (48:58)… Airwallex is a leading global payments and financial platform for modern businesses, offering trusted solutions to manage everything from business accounts to… Graham Allison on global realignments: Iran, China, Israel, Greenland.

  • EN Highlights:
    • (0:00) The Besties welcome legendary Harvard professor Graham Allison (1:14) Iran Conflict: Strategy, Netanyahu’s influence, Trump’s motivation, redef…
    • (24:28) China: Taiwan invasion, trade, and Trump’s April meeting (39:50) Greenland: Importance, deal scenarios, EU socialism angle (48:58)…
    • Airwallex is a leading global payments and financial platform for modern businesses,offering trusted solutions to manage everything from business accoun…

Lenny’s Podcast (A_full) Link to heading

• How I built a 1M+ subscriber newsletter and top 10 tech podcast | Lenny Rachitsky

  • Published: 2026-03-12 20:32 Beijing Time
  • Abstract: - People have been asking me to sit on the other side of the mic for a long time.
    • With my wife’s debut children’s book, Charts for Babies, coming out next month, we figured: why not do it together?
    • What followed was one of the most honest conversations I’ve had on this podcast.
    • Michelle asked things no one else would think to ask—and many things I’ve never shared publicly.
    • You’ll hear about the specific moment that pushed me to start my newsletter, my philosophy on quality and iteration, what stresses me out the most, and the scariest moment of my life.
  • EN Highlights:
    • People have been asking me to sit on the other side of the mic for a long time
    • With my wife’s debut children’s book, Charts for Babies , coming out next month, we figured: why not do it together
    • What followed was one of the most honest conversations I’ve had on this podcast
    • Michelle asked things no one else would think to ask—and many things I’ve never shared publicly

• From Figma to Claude Code and back | Gui Seiz & Alex Kern (Figma)

  • Published: 2026-03-11 20:03 Beijing Time
  • Abstract: - Most teams are still passing static design files back and forth, and most Figma files are already out of date by the time they reach engineering.
    • Figma’s Gui Seiz (designer) and Alex Kern (engineer) present on screen the exact workflow their team uses to bridge the gap with AI.
    • They demonstrate how to use Figma MCP to pull a running web app directly into Figma, make collaborative edits, and then push it back into code.
    • Replacing it is a fluid, two-way loop where design and code inform each other in real-time.

• Listen or watch on YouTube, Spotify, or Apple Podcasts What you’ll learn: - How to use Figma’s MCP to pull production code directly into a Figma file - The workflow for pushing design changes from Figma back to the codebase with Claude Code, without manual CSS adjustments - How to export multiple code states (e.g., all five states of a registration flow) into Figma so designers can work with what actually exists in production - Why AI is shifting design work from upstream planning to downstream craftsmanship, eliminating the rushed execution phase in the middle - How to create custom skills that automate pre-flight checks, lint fixes, and CI monitoring before shipping code to production - How to structure your codebase so that AI can more effectively write 90% of it for you Brought to you by: Optimizely — AI agent orchestration platform for marketing and digital teams In this episode, we’ll cover: ( 00:00 ) Intro to Gui and Alex from Figma ( 02:56 ) How AI is changing Figma’s internal workflow ( 05:17 ) The breakdown of the linear design-to-code workflow ( 07:28 ) Demo: Pulling production code into Figma with MCP ( 10:49 ) Using Figma for precise design manipulation and team collaboration ( 14:10 ) Demo: Pushing Figma designs back to code with Claude Code ( 16:06 ) How AI is changing the role of a software engineer ( 18:43 ) The shift to upstream planning and downstream craftsmanship (22:31) Demo: Exporting multiple code states back to Figma (25:23) Sync vs. async

  • EN Highlights:
    • Most teams are still passing static design files back and forth, a…
    • Gui Seiz (designer) and Alex Kern (engineer) from Figma walk through the exact workflow their team uses to bridge that gap wit…
    • They demo how to pull a running web app directly into Figma using the Figma MCP, edit it collaboratively, and push it back to code
    • The old linear waterfall workflow is gone

• A guide to advanced B2B positioning

  • Release time: 2026-03-10 20:45 Beijing time
  • Abstract: - Every week, I answer reader questions about building products, driving growth, and accelerating careers.
    • As AI makes it trivial to build and launch products (and soon even to come up with product ideas), the biggest challenge for product teams is quickly becoming distribution: getting people to pay attention to your product amidst the ever-increasing noise of launches.
    • One of the most powerful tools for cutting through the noise is positioning.
    • Strong, specific positioning grabs attention and helps people instantly understand why your product is for them.
    • April Dunford is the world’s leading expert on the subject.
  • EN Highlights:
    • 👋 Hey there, I’m Lenny
    • Each week, I answer reader questions about building product, driving growth, and accelerating your career
    • For more: Lenny’s Podcast | Lennybot | <a href=“https://www…
    • via=lenny”>public speaking course , and interview prep copilot <a class…

• A guide to advanced B2B positioning

  • Release time: 2026-03-10 17:00 Beijing time
  • Abstract: - April Dunford, the world’s leading expert on B2B positioning and author of “Beyond Awesome,” has spent 10 years working with over 300 companies to help them stand out.
    • In this episode, she goes beyond the basics to share strategies for overcoming the four trickiest positioning hurdles teams face.

• Now Listening: YouTube | Apple | In this episode, you will learn: - The four obstacles every team faces when making positioning decisions - Why cross-functional teams almost always disagree on who they are competing against - How to break the cycle of product pessimism - Why the “So what?” test is crucial for defining differentiated value - How to use a GTM strategy to drive positioning decisions References: - April Dunford: - April Dunf… Learn more 10 Years, 300 Companies, 4 Non-Obvious Lessons for Overcoming the Toughest Obstacles Obviously Awesome, spent 10 years working with 300+ companies… Advanced B2B Positioning Guide.

  • EN Key Points:
    • <a class=“image-link image2 is-viewable-img” href=“https://substackcdn.com/image/fetch/$s_
    • ,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F46ee1b7a-be53-45bc-bd74-a548414a8b44_3016x3016.p…
    • ,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F46ee1b7a-be53-45bc-bd74-a548414a8…
    • April Dunford, the world’s leading expert on B2B positioning and author of Obviously Awesome, has spent 10 years working with more than 300 compa…

• 🎙️ This week on How I AI: Mastering Midjourney: How to create consistent, beautiful brand imagery without complex prompts

  • Published: 2026-03-09 23:03 Beijing Time
  • Summary: - Every Monday, host Claire Vo shares a 30-to-45-minute episode with a new guest, showcasing practical and effective ways they’ve learned to use AI in their work or life.
    • No pontificating—just specific and actionable advice.
    • Mastering Midjourney: How to create consistent, beautiful brand imagery without complex prompts Provided by: - Vanta — Automate compliance and simplify security - Lovable — Build apps simply by chatting with AI Jamey Gannon is an AI creative director who has spent thousands of hours working in Midjourney.
    • In this episode, she will introduce her system for generating consistent, high-quality brand imagery using mood boards, style references, and personalized code.
    • She and Claire cover the complete workflow for transforming loose inspiration into a scalable visual system, how to avoid the biggest mistakes people make when using AI image tools, and why a single image reference can often outperform a paragraph of prompts.
  • EN Key Points:
    • <a class=“image-link image2” href=“https://substackcdn.com/image/fetch/$s_
    • ,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F361d81ef-7faf-4d8e-8028-5d5e03432a9a_2329x551.pn…
    • ,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F361d81ef-7faf-4d8e-8028-5d5e03432…
    • No pontificating—just specific and actionable advice.

• Mastering Midjourney: How to create consistent, beautiful brand imagery without complex prompts

  • Publication Time: 2026-03-09 20:03 Beijing Time
  • Summary: - Jamey Gannon is an AI creative director who excels at creating consistent and beautiful brand imagery using AI tools.
    • In this episode, Jamey demonstrates her streamlined workflow for generating cohesive brand assets using Midjourney, Nano Banana, and other AI image tools.
    • She details her process of creating mood boards, using style references, developing personalization codes, and strategically iterating to achieve a consistent aesthetic.
    • Jamey shows how visual references and strategic shortcuts can produce better results with less effort, rather than relying on complex prompts.
    • Listen or watch on YouTube, Spotify, or Apple Podcasts What you will learn: - How to create effective mood boards that communicate your desired aesthetic to AI image generation tools - Why style references (SREF) often produce more consistent results than a general mood board in Midjourney - A systematic approach to testing and refining a visual style - How to use personalization codes in Midjourney to develop your own unique aesthetic preferences - Techniques for combining image references, style references, and minimal prompts to achieve consistent brand imagery - A workflow for using Nano Banana to fix specific elements in Midjourney-generated images without extensive editing - How to package and deliver your brand imagery system to clients so they can continue generating consistent assets Brought to you by: Vanta — Automate compliance and streamline security Lovable — Build apps by simply chatting with an AI In this episode, we’ll cover: ( 00:00 ) Introduction to Jamey Gannon ( 02:31 ) Creating a mood board as the foundation for AI image generation ( 08:45 ) Using SREF for better consistency ( 11:15 ) Test prompts for evaluating style consistency ( 12:33 ) The iterative process of creating and refining images ( 24:28 ) Combining techniques for consistent brand imagery ( 28:25 ) Extending your aesthetic across different subjects ( 35:48 ) Using Nano Banana for targeted image refinement ( 38:23 ) Creating a realistic AI self-portrait ( 43:04 ) Building a visual reference library for inspiration ( 46:50 ) Troubleshooting tips for when AI doesn’t cooperate Referenced Tools: • Midjourney: • Nano Banana: • Flora: • Pinterest: • Cosmos: Other References: • Style References (SREF) in Midjourney: Where to find Jamey Gannon : Website: LinkedIn: X: Instagram: Maven Course (10% off with this link): Where to find Claire Vo: ChatPRD: Website: LinkedIn: X: Production and marketing.
  • EN Key Points:
    • Jamey Gannon is an AI creative director who speci…
    • In this episode, Jamey demonstrates her streamlined workflow for generating cohesive brand assets using Midjourney, Nano Banana, and other AI image tools
    • She walks through her process of creating mood boards, using style references, developing personalization codes, and strategically iterating to achieve a consis…
    • Rather than relying on complex prompts, Jamey shows how visual references and strategic shortcuts can produce better results with less effort.