THE AI UPSKILL
A weekly digest for Beginner Builders
⚡ TLDR
Nvidia just had the biggest quarter of any company in tech history with $68 billion in three months, driven almost entirely by AI. The numbers confirm what many people still aren't ready to accept: this isn't hype anymore, it's infrastructure. Meanwhile, every major AI CEO flew to India for a summit that showed just how fast the rest of the world is catching up.
This week we're also showing you how to build a simple AI agent that reads your messages and sends emails on your behalf. No coding required. And if you've never tried NotebookLM, we've got a five-minute setup that turns any document into a podcast you can listen to on your commute.
ARTIFICIAL INTELLIGENCE
THIS WEEK WE'RE WATCHING
→ A guide to Gemini Canvas: Did you know you can connect NotebookLM to Gemini Canvas and spin up a website in seconds? Think about it, take your research or dump a few videos into Notebook and after a few clicks you have a functioning website. And that’s just scraping the surface on what it can do. Check out Paul Lipsky’s video here.
→ The clearest breakdown of Claude's three modes: Use Claude Chat for thinking things through, CoWork for getting things done end-to-end, and Claude Code for building. Simple in theory, but knowing which to reach for saves a lot of wasted time. This overview by Grace Leung nails it. Watch it here.
Want to go deeper on any of this? Join the AI-Enabled Marketer Skool community, which takes you from absolute beginner to AI-Enabled Marketer with Myosin.xyz, a community of top marketing professionals. https://www.skool.com/the-ai-enabled-marketer/classroom
THE BUILDERS TIP
💡 AI PLAYGROUND: TURN ANY DOCUMENT INTO A PODCAST WITH NOTEBOOKLM
Reading 30 pages of research notes is painful. Listening to two AI hosts break it down for you while you commute? That's NotebookLM.
Google's NotebookLM (notebooklm.google.com) takes your documents — PDFs, Google Docs, slides, even YouTube links — and turns them into things you'd actually use: AI podcasts, flashcards, quizzes, mind maps, study guides, and now even slide decks, infographics and data tables.
The killer feature is Audio Overviews.
Upload your sources, click generate, and you get two AI hosts who banter, explain, and break down your content like a real podcast. It's not robotic text-to-speech — it sounds like two people having a genuine conversation about your material.
Try this right now (it's free):
Go to notebooklm.google.com
Create a new notebook
Upload a PDF, paste a YouTube link, or connect a Google Doc
Click "Audio Overview" in the Studio panel on the right
Hit Generate
What makes it powerful:
You can customize the podcast — tell it to focus on a specific topic, adjust complexity, or change the length
Interactive Mode lets you interrupt the hosts mid-conversation to ask follow-up questions
Everything stays grounded in YOUR sources — no hallucinations from random internet data
Your uploads aren't used to train Google's models
Example prompts to try:
Upload a competitor's whitepaper → "Generate an audio overview focusing on their pricing strategy"
Paste 3 YouTube tutorial links → "Create flashcards covering the key techniques from all three videos"
Upload meeting notes → "Generate a quiz to test if my team understood the key decisions"
LET”S BUILD
SIMPLE BUILD: EMAILING AI AGENT
What if you could send emails by just... telling AI what to say?
This week's build is a chat-based email assistant. You type natural language into a chat window — "Send an email to John about rescheduling our meeting to Friday" — and the AI drafts it, looks up the contact, and sends it through Gmail.
How it works: You chat with an AI agent → it checks your contacts Google Sheet for the recipient's email → drafts a professional email → sends it via Gmail. It remembers your last 10 messages, so you can say things like "now send that same update to Sarah" and it knows what you mean.
QUICK START:
Grab the workflow JSON from our Skool community here and import it into n8n
Create a Google Sheet called "Contacts" with two columns: Name and Email. Add a few contacts.
Connect your credentials:
OpenAI — on the OpenAI Chat Model node
Gmail — on the Gmail Tool node
Google Sheets — on the Contacts node (select your Contacts sheet)
Open the AI Agent node → in the system prompt, replace [YOUR NAME] with your actual name (this is how the AI signs off your emails)
Click "Chat" at the bottom of the canvas to open the chat window
Test it: Type: "Send an email to [contact name] letting them know I'll be 15 minutes late to our meeting."
Check your Gmail sent folder — it should be there.
Something not working? Copy the chat input and agent output, then paste it into Claude or ChatGPT for troubleshooting.
MAKE IT YOURS:
Change the tone — The system prompt says "professional and neutral." Want it casual? Friendly? Direct? Edit the system prompt in the AI Agent node to match how you actually write.
Expand your contacts sheet — Add columns for company, role, or notes. The AI can use that context: "Email the marketing lead at Acme about our proposal" becomes possible when your sheet has that info.
Swap the model — We use gpt-4o-mini for speed and cost. For more nuanced emails, try gpt-4o or Claude Sonnet.
IN OTHER NEWS
THE BIGGER PICTURE
Nvidia Just Reported the Biggest Quarter in Tech History.
On Feb 25, Nvidia reported quarterly revenue of $68.1 billion, up 73% from the same time last year. Its data centre business alone brought in $62.3 billion, driven almost entirely by AI chip demand. Net income nearly doubled to $43 billion. Nvidia is now the world's most valuable company, with a market cap exceeding $4.7 trillion.
Not everyone is celebrating. Critics point out that roughly 70% of Nvidia's revenue comes from just eight companies, and nobody has yet proven those companies are making enough money from AI to justify the bill. Microsoft, Amazon, Google and Meta are on track to spend $700 billion on AI infrastructure in 2026. The returns on that investment remain an open question.
Our take: Every dollar of that $68 billion is companies paying to run AI at scale. The infrastructure wave is no longer a promise, it's a quarterly earnings report. The skeptics raise fair points about ROI, but the companies writing those cheques have access to the same data we do. If you're still deciding whether to take AI seriously in your work, that window is closing.
(Sources: CNBC, TechCrunch, Fortune)
🔥 THE WATER COOLER:
Every Major AI CEO Just Flew to India. Here's What Happened.
Last week, India hosted a four-day AI Impact Summit that drew Sam Altman, Dario Amodei, Sundar Pichai, and Demis Hassabis, alongside Prime Minister Modi and French President Macron.
The headline numbers were staggering. India announced a $1.1 billion state-backed AI venture fund. The Adani Group committed $100 billion to build renewable-powered AI data centres by 2035. Altman revealed India now has over 100 million weekly ChatGPT users, second only to the US, with 18 to 24 year olds driving nearly half of all usage. OpenAI also confirmed it will open two new offices in Bengaluru and Mumbai.
Our take: AI adoption is no longer a Silicon Valley story. It's a global one, and India just made its ambitions clear. A country of 1.4 billion people, half under 30, is going all in. That changes the scale of everything that comes next.
(Source: TechCrunch)
That's it for this week.
Building something cool with AI? Reply to this email. I feature reader projects. 💬
P.S. If this was useful, forward it to someone upskilling in AI.