Exciting AI Updates Weekly - May 1, 2026

Exciting AI Updates Weekly - May 1, 2026 Presented by Lev Selector Slides on GDrive: https://drive.google.com/drive/folders/1dmj00iilSP-JYmriMUUaHCKz5X54e3UX?ths=true Slides on GitHub - https://github.com/lselector/seminar (click on pptx file, then on "raw" or download button on the right) ------------------------------------------------------- Question to you: Do you use a personal Digital "Employee"? If not, then why not? It is so easy to start! ------------------------------------------------------- Need AI Consulting ? - Enterprise AI Solutions - https://EAIS.ai - Linkedin - https://www.linkedin.com/in/levselector - GitHub - https://github.com/lselector ------------------------------------------------------- Our books: "Artificial Intelligence In Business: Strategies That Work" https://www.amazon.com/dp/B0FKZ3364G/ "AI Product Manager: Building Tomorrow" https://www.amazon.com/dp/B0FMS9NFK9/ "How to Learn with AI?: A Practical Guide for Students, Professionals, and Self-Learners" https://www.amazon.com/dp/B0FWLNVQMJ/ "Professional + AI: Future-Proofing Your Career" https://www.amazon.com/dp/B0GBXLCTZQ "FP&A + AI: How to Work Faster, Smarter, and Deliver More Value" https://www.amazon.com/dp/B0FY6Z5BBW --------- Contents of today's video: 00:00 Introduction 00:20 Plan first, Code second, Verify always - Boris Cherny 00:25 Crowd-sourced "LM Arena" Leaderboard 01:10 DeepSeek V4 Pro 02:24 Grok 4.3 Beta 02:59 Chinese Models 03:48 Agentic Harness For Different Models 05:06 Context Length 07:17 GenericAgent - minimal self-evolving 07:45 Claude desktop app Improvements 08:23 Hermes Agent 08:49 Problems with OpenClaw 09:38 How to Survive AI 10:28 Google's Hope - Continual Learning POC 11:07 Huashu Design - Open Source Skills 11:46 Boris Cherny 6-step Claude Code System 12:26 Claude Code - 15 Rules & Workflows 12:56 Personal Agent OS 13:44 Google Gemini Deep Research Max 14:16 How to reduce token usage 15:20 News Research Workflow 15:51 Telegram Lobster Father 16:28 OpenClaw Updates 16:44 China is Winning the AI Race 18:51 From Transformers to Recurrent Models 19:54 AI Interview Question 20:54 AI Consulting - From Projects to Retainers 22:13 Anijam AI - Animation Agent 22:41 Dan Kennedy Book 23:28 About Job Search in the US 25:20 Jobs & Layoffs ---- Takeaways: 1. China is rapidly winning the open-source AI race with massive government-backed models like DeepSeek V4 Pro (1.6T parameters), while US labs increasingly close their models, creating a strategic risk where US businesses end up depending on Chinese AI standards and hardware. 2. Every serious AI model now needs a harness - Claude Code, Codex, Gemini CLI - because no model can be trusted off-leash, and tools like the personal agent OS (identity, context, skills, memory, connections, verification, automation) are becoming the foundational layer for portable, reusable AI setups. 3. AI gives consultants huge leverage - one person can serve four clients on 5,000 per month retainers because the human becomes faster than the client, but survival in the AI economy requires entrepreneurial mindset, authentic human connection, AI proficiency, and building a personal brand. ---- Quotes: "Plan first, code second, verify always." "We're losing the AI race to the country we sold the chips to." "You validate AI output the same way you validate a toddler's homework - carefully and with low expectation." ---- Summary: In this weekly AI update from May 1, 2026, I cover the latest leaderboard with DeepSeek V4 Pro and Grok 4.3 Beta, the dominance of Chinese open-source models, agentic harnesses for every major model, and growing context windows. I review Claude Code improvements including Boris Cherny's six-step system and fifteen practical rules, the Hermes browser agent, problems with OpenClaw, Google Hope continual learning, the Huashu open-source design skills, and the personal agent OS concept. I discuss how to reduce token usage, news research workflows, the Telegram Lobster Father agent builder, why China is winning the AI race, the shift from transformers to recurrent models, validating AI output, and how AI consultants can stack retainers. I close with Dan Kennedy's new book, advice on US job search avoiding Dice and Monster, and the week's layoffs at Meta, Microsoft, Snap, and Nike totaling around 18,000 jobs. ---- Keywords: DeepSeek V4 Pro, Grok 4.3, Chinese models, open source, agentic harness, Claude Code, Boris Cherny, context window, GenericAgent, Hermes agent, OpenClaw, Google Hope, continual learning, Huashu design, personal agent OS, Gemini deep research, token usage, news research, Telegram Lobster Father, transformers, recurrent models, AI validation, AI consulting, retainers, Anijam AI, Dan Kennedy, job search, layoffs, LM Arena, mixture of experts ---- ---- ----