Welcome back to the AI Edge weekly newsletter!
In this week’s edition, we’ve put together a full free 30-day AI mastery and learning course.
This course is divided into four weeks, each with a distinct focus, and culminates in getting started with your AI automation journey.
Week One: Foundational AI Skills
Week Two: Tool Selection & Experimentation
Week Three: Leveraging AI for Real Work
Week Four: Advanced Workflows & Hermes Automation
By the time you’re done reading, you’ll have an exact 30-day AI learning roadmap that you can implement to start automating your life.
After our main anchor piece, be sure to stick around for the Midweek Edge for all the latest AI news updates, and our Looking Ahead section, where we compile our top AI research.
Let’s get right into things!
What You'll Need
Before you start, here's what this course requires:
~60 minutes a day, distraction-free (more if you can spare it)
A desktop or laptop + WiFi
A Google account for quick sign-ups across AI tools
Curiosity and discipline to actually follow through
Optional: budget for paid AI subscriptions ($20-$100/month range)
Week One: Foundational AI Skills
Start here: Create a new ChatGPT project called "30-Day AI Learning Roadmap."
In Settings, set memory to Project-only so this stays isolated from your other chats:
In your project context, go ahead and paste this:
"I am a beginner learning AI over the next 30 days.
I will use this project to document my progress, store my best
prompts, and get personalised guidance.
My roadmap:
Week One — Foundational Skills.
Week Two — Tool Experimentation.
Week Three — Real Work.
Week Four — Advanced Workflows.
Always remember where I am in my journey and tailor your
responses to my current skill level.
My goal by Day 30 is a fully functioning personal AI workflow."For the first seven days, you'll master three core AI skills:
Prompt engineering
Model stacking
Task delegation
To get started with your Week One learning, start a new chat and paste this prompt:
"I am on Day 1 of my 30-day AI learning journey.
My goal for Week One is to master prompt engineering,
model stacking, and task delegation over the next 7 days.
Act as my personal AI coach. Each day, give me one focused
lesson per skill, three hands-on exercises I can complete
in under 60 minutes, and 10 reflection questions.
Start with Day 1 and do not move forward until I tell you
I'm ready. I am a beginner."Every day for the next seven days, dedicate at least 60 minutes to learning with ChatGPT.
It will teach you the basics of prompt engineering, model stacking, task delegation, and other essential AI fundamentals through personalized tutoring.
After Day 7, you’ll take an AI fundamentals test generated by ChatGPT.
It will test your current knowledge, and you’ll need a score of 80+ to progress to Week Two.
"I have completed all 7 days of Week One.
Give me a comprehensive but fair assessment:
a 25-question knowledge test covering all three skills, a
practical exercise for each skill I must complete in
real time, and a final challenge combining all three.
Grade me out of 100. I need 80+ to progress.
If I score below 80, tell me exactly what to review and
retest me on those areas only."Week Two: Tool Selection & Experimentation
This week is all about getting your hands dirty with real AI tools.
To kick this week off, go ahead and create free accounts with these five AI tools:
Claude: HTML coding, writing, file analysis
Gemini: Google Workspace integrations
Manus AI: Spreadsheet work, basic automations, email
Notion: Automated CRMs, agents, connections
NotebookLM: AI tutor, research, learning
Then, go back to your ChatGPT project and paste this:
"I've completed Week One and passed my foundational skills test.
I'm entering Week Two: Tool Selection & Experimentation
with Claude, Gemini, Manus AI, Notion, and NotebookLM.
Before we begin, ask me questions one at a time to understand
how I work. Once you have a full picture,
build me a 7-day experimentation plan: one tool per day, a
specific mini-project for each, 3 prompts to get the most out
of each tool, and a comparison question at the end of each day.
At day 7, give me a Tool Selection Matrix scoring each tool on ease of
use, relevance, time saved, and output quality.
Do not move forward each day until I tell you I'm ready."This prompt will help you build a 7-day AI tool experimentation plan to test these AI tools throughout Week Two.
There is no right or wrong way to spend your time here, as long as you're actively getting your hands dirty & testing the 5 foundational AI tools above, you're golden.
Think of this entire week like being in a Playground Phase.
Weeks 3-4 are dedicated to leveraging AI to complete real, autonomous work (don't worry if you're not yet completing tasks autonomously; the goal is just to experiment here).
Week Three: Leveraging AI for Real Work
You've tested the tools. Now it's time to actually use them for real work.
Before automating anything, you need data on how you actually spend your day.
On day one of Week Three (Day 15/30 total), you'll conduct the Time Audit.
The goal of this exercise is to log everything you do over a full day and collect data on where you can start leveraging AI in real work.
To execute this, go back to your 30-day AI Learning project inside ChatGPT, and send in this prompt:
I have completed Weeks One and Two of my 30-day AI learning roadmap. I now have a foundational understanding of prompt engineering, model stacking, and task delegation, and I have experimented with several AI tools.
I am now entering Week Three: Leveraging AI for Real Work.
To kick off this week, I want to do a full Time Audit. Here is everything I did today (paste your full day here):
Based on what I've written, I want you to:
1. Categorize every task into one of three buckets: tasks AI can fully handle, tasks AI can assist with, and tasks that require me fully
2. Rank the top 5 tasks where using AI would save me the most time or produce the best results
3. For each of those 5 tasks, suggest the exact AI tool and a starter prompt I can use right away
4. Give me a simple daily workflow I can follow this week that integrates AI into my real work routine
Be specific. Use what you know about my skill level and the tools I've already experimented with.In this prompt, you’ll want to include your full-time log sheet.
ChatGPT will then return the exact workflows that you should build with the five AI tools you signed up for earlier.
Some example workflows to automate (based on your time audit):
Example: You spend 30 minutes/day on Slack & emails
Solution: Manus + Slack AI connectors for summaries/thread management
Example: You spend 60 minutes/day on Excel & spreadsheet work
Solution: Claude in Excel or Gemini for Sheets
Example: You spend 1-2 hours/week creating presentations
Solution: GammaAI or Claude + Excalidraw MCP
By the end of Week Three, you should have experimented with building 7 AI workflows (one each day).
Week Four: Advanced Workflows & Hermes Automation
By now, you've built 7 beginner AI workflows. Some worked, and some probably didn't.
This final week is all about doubling down on what worked and making it run without you using Hermes agent.
Paste your 7 workflows into a new chat in your project, noting which ones succeeded and which didn't:
"I've completed Weeks One through Three.
I've built 7 workflows — here they are, noting
which worked and which didn't: [paste your workflows].
I'm entering Week Four: Advanced Workflows.
The goal this week is to use Hermes agent to double down
on the workflows that worked well.
Before we begin, ask me questions one at a time to
understand why certain workflows succeeded and others failed.
Then: identify the 2-3 workflows with the most potential and
explain why. Then, show me exactly how to build each one with Hermes.
My goal by Day 30 is a lean, reliable personal AI
system that saves meaningful time every day."
After sending this prompt, your ChatGPT project will suggest and walk you through exactly how to use Hermes to automate your Week Three workflows that were actually high-ROI.
This could be:
Email management
Before: Manus + Gmail connector.
Now: Hermes scans, summarises, and responds on your behalf autonomously.
CRM Management
Before: Notion CRM with NotionAI for morning briefs.
Now: Hermes creates a daily morning brief report and pings you on WhatsApp daily.
Daily Research
Before: SuperGrok for deep research.
Now: Hermes connects via the X MCP to produce an autonomous daily research report.
By the end of Week Four, your goal is to have 2-3 fully autonomous Hermes workflows based around your real needs.
Quick Recap
Week One: Master the foundational AI skills with ChatGPT (prompt engineering, model stacking, task delegation).
Week Two: Sign up and experiment with 5 tools: Claude, Gemini, Manus AI, Notion, NotebookLM.
Week Three: Run a Time Audit, and build 7 workflows to solve real pain points
Week Four: Refine your best workflows and turn them into fully autonomous systems with Hermes
Foundational skills → light experimentation → real work → autonomous systems.
Follow this roadmap, and by Day 30, you'll have a lean, reliable personal AI system most people will never build.
Midweek Edge
Our manually curated list of the most important AI news updates:
Google Omni & Nano Banana 2 Lite preview
Google just released Google Omni and Nano Banana 2 Lite in preview.
Omni: Multimodal LLM that can generate full videos with audio
Nano Banana 2 Lite: Most cost-efficient image model
To try these new models:
Head to Google AI Studio: https://aistudio.google.com/prompts/new_chat
Swap your model selection to “Omni” or “Nano Banana 2 Lite”
𝕏AI Launches Official MCP
𝕏AI just announced the hosted 𝕏 MCP.
Your AI agents now have direct access to 𝕏 data.
Connect directly to Grok Build, Cursor, Claude Code & more:
Anthropic Re-Releases Fable 5
Fable 5 is officially back! Anthropic has come to an agreement with the US government.
Starting today, you can access Fable 5 in your Claude plan (it'll be live in a few hours).
Everything you need to know about Fable 5 re-release:
→ The original jailbreak that got Fable banned was tested against every major competitor and found that they could produce the exact same outputs - not a catastrophic case
→ Anthropic trained a new safety classifier that blocks the specific jailbreak technique in 99%+ of cases
→ Trade-off: it now flags more general coding and debugging requests too
→ Mythos 5 still restricted, restored only for approved US orgs via Project Glasswing
→ Pro, Max, Team, and Enterprise plans get Fable 5 included for up to 50% of weekly usage limits through July 7, then shifts to API
Anthropic Announces Sonnet 5
Sonnet 5 is finally live, here’s the TL;DR:
→ It's the most agentic Sonnet model Anthropic has ever shipped, closing the gap with Opus 4.8 while staying significantly cheaper.
→ Default model for Free and Pro plans, available everywhere now
→ Launch pricing: $2/$10 per million tokens (input/output) through August 31, then $3/$15 after
→ Lower hallucination and sycophancy rates than Sonnet 4.6
→ Better at refusing malicious requests and resisting prompt injection attacks
→ Available in Claude Code, Claude Platform, and via API as claude-sonnet-5
Anthropic Announces Claude Science
Anthropic just announced Claude Science - a new AI platform built for scientific research (like Claude Design).
→ Available in all paid plans now
→ Brings together researcher tools (like PubMed) into Claude
→ Connects to 60+ scientific databases
How to get started
Download the Claude Science app and log in: https://claude.com/product/claude-science
OpenAI Announces GPT-5.6 Family
OpenAI has introduced a limited preview of the GPT-5.6 model family.
GPT-5.6 Sol: Frontier model
GPT-5.6 Terra: Balanced model
GPT-5.6 Luna: Fast, affordable model
Unfortunately, the US government has restricted the public release of these models, but here is how the latest OpenAI frontier model lines up against Mythos:
@chrissgpt on X
Apple Products Price Increases
Last week, Apple announced 10%- 50%+ price increases across iPads, Macs, and more due to memory chip costs.
Looking Ahead
Our manually curated list of the top AI trends, research, workflows & more.
Robotics Research
There has been a lot of discussion around the robotics industry lately.
Venture capital investments just hit a quarterly high at $16B+.
We did some research and came up with a curated list of the top robotics companies for potential investments (nfa):
Taste Skill
This agentic skill is based on thousands of Reddit comments and removes AI slop from UI/UX design.
It kills all signs of AI-generated slop - blocky logos, fonts, etc.
→ https://github.com/JCarterJohnson/vibecoded-design-tells
Reducing AI Costs
Coinbase recently announced that it reduced its AI token spend by ~40% while increasing its total token usage.
Here's how (save these tips):
→ Default to open-source models
→ Automated model routing
→ Warm cache when possible
→ Keep context lean
→ Full org-wide token usage visibility
Closing out
If you made it this far, thank you for reading, and we hope you found this week’s edition valuable.
If you enjoy reading, please share this with someone who you think would benefit from our publication.💙
Our promise to you: Every Wednesday, at 7 am EST, we’ll cut through the AI noise and send you human-curated AI content to make sure you stay ahead of the AI curve (100% free always).
Content Pipeline
YouTube
Live Now: GLM-5.2 Guide
X (Twitter)
Live Now: Building Agentic Skills for Claude
Free AI assets
For all the AI prompts, cheatsheets, and PDF guides mentioned on the AI Edge YouTube channel, grab them by joining Miles’ free Instagram community here: