Hey everyone! 👋

This week continues our deep dive into how to communicate clearly with AI. In AI Education, we’ll simplify the art of prompting—why clear structure beats clever wording and how a five-piece framework keeps your asks sharp and on-target.

Our 10-Minute Win helps you lock in free money by building a 401(k) Match Maximizer—a quick, high-ROI workflow every reader can use today.

And in Founder’s Corner, I’m sharing my perspective on the “AI is killing jobs” headlines—and how I filter fear-driven stories to focus on what actually matters for my career and learning path.

Thanks for being here and staying committed to learning, one focused week at a time.

1) Bloomberg: Apple to use Google’s AI model to run the new Siri

Summary: Apple is near a deal to power a revamped Siri with Google’s Gemini, reportedly paying around $1B/year while Apple continues building its own models.

Why it matters: If Apple leans on Google, Siri upgrades land faster—and the Big Three model providers become even more intertwined across the stack.

2) Microsoft signs a $9.7B deal with IREN to secure Nvidia GB300 chips

Summary: A five-year contract gives Microsoft dedicated GB300 capacity at IREN’s 750-MW Texas campus (liquid-cooled builds), easing near-term compute bottlenecks.

Why it matters: Pre-buying third-party capacity lets Microsoft scale Copilot and cloud AI without waiting for new MS-owned data centers.

3) OpenAI inks a 7-year, $38B cloud deal with AWS

Summary: OpenAI will tap AWS for hundreds of thousands of Nvidia GPUs over seven years, broadening beyond Microsoft’s cloud after its restructuring.

Why it matters: Vendor diversification reduces supply risk and could speed roadmaps for ChatGPT, Sora, and agents—while reinforcing AWS as a core AI infra provider.

4) Meta plans a $600B U.S. investment push for AI data centers

Summary: Meta outlined a three-year plan centered on massive U.S. data-center buildouts and financing deals to scale compute for next-gen AI.

Why it matters: The AI build-out is now about power, cooling, and real estate at unprecedented scale—second-order winners include construction, utilities, storage, and interconnects.

5) Altman urges expanding CHIPS Act tax credits for AI growth

Summary: OpenAI CEO Sam Altman pressed U.S. officials to extend CHIPS Act tax incentives to AI data centers and related infrastructure, arguing it’s critical for competitiveness.

Why it matters: If policy tilts toward AI infra, expect faster capacity build-outs—and clearer economics—for the whole stack (chips, power, data centers).

Over the last few issues, you learned three big ideas: how AI represents meaning as number lists (vectors), how it reads text as small pieces (tokens), and that it reads on a fixed-size page (the page limit). Put those together and you get the rule that drives everything: the model does its best work when you give it the right meaning, in the right pieces, that fit on the page.

This week is about how you ask. A good prompt is not a trick—it’s clear directions plus only the facts that matter. You name the goal, share the key info, say the format you want back, set simple limits, and (optionally) show a tiny example. That structure gives the model a pattern to follow, keeps your message inside the page, and helps the model focus on the meaning that matters.

We’ll use the same personal finance tasks you already know—bank statements, bill emails, and budget notes—to practice this. By the end, you’ll know how to shape your ask so the model reads what it needs, understands the point, and answers cleanly.

Core lesson — definition first, then a simple 5-piece structure

What is prompt engineering:Prompt engineering is the practice of designing your ask so the model can read the right information, understand the point, and answer in the form you need. It means choosing only relevant facts, giving clear instructions, using simple structure (titles, bullets, examples), and ordering things so the key parts fit on the model’s page.

How this ties to what you learned:

• Models turn text into meaning (those number lists).
• They read small pieces of text on a fixed-size page.
• Good prompting shapes the meaning, pieces, and order so the right text fits—and the answer stays on-target.

The 5-piece structure:

Use this for almost everything you ask:

1) Goal - Say exactly what you want. One sentence.

2) Key info - Give only the facts the model needs. Keep it short and relevant.

3) Format - Say how you want the answer. For example: bullets, a short plan, or a “yes/no plus a quoted line.”

4) Constraints - Set limits so the answer stays tight. For example: “Three bullets.” “Focus on grocery spend only.” “No extras.”

5) Tiny example (optional) - One short example shows the pattern you want. Keep it small.

Why this works: The model follows patterns in your words. A clear pattern helps it stay on topic, fit the page, and answer in the form you asked for.

Contrast & clarity — vague vs. clear

Vague: “Look at my statement and tell me what you think.”

Clear (uses the 5 pieces):

• Goal: “Find unusual spending this month.”
• Key info: “Use the lines below from my statement.”
• Format: “Return three bullets: merchant, date, amount.”
• Constraints: “Only items over $100. No extras.”
• Tiny example: “Example: ‘ACME Grocery — 10/12 — $128.42’.”

The clear version gives the model a pattern to copy and a tight scope to follow.

Examples that land — one monthly money thread

Example 1 — Bank statement check (unusual spending):

• Prompt:
  • Goal: Find unusual spending this month.
  • Key info: Use the statement lines below.
  • Format: Three bullets: merchant, date, amount.
  • Constraints: Only over $100. No commentary.
  • Tiny example: Example: “ACME Grocery — 10/12 — $128.42”.
  • Lines: (paste 10–20 relevant lines only, not the whole month)

Example 2 — Bill credit check (was the credit applied?):

• Prompt:
  • Goal: Confirm if a promised $25 credit was applied on the October bill.
  • Key info:
    • Promise date: Sept 15
    • Amount: $25
    • Account ending: 4931
    • Bill lines: (paste the section that shows credits or adjustments)
  • Format: “Yes” or “No” + the exact line that proves it.
  • Constraints: If “No,” give the most likely reason in one sentence.

Example 3 — Budget plan (cut $100 next month):

• Prompt:
  • Goal: Create a plan to cut $100 from next month’s spending.
  • Key info:
    • This month: groceries up $120 (party), restaurants down $75, two new ride shares
  • Format: Three steps, one sentence each.
  • Constraints: No guilt, no shaming. Keep steps practical.
  • Tiny example: Example step: “Swap one restaurant meal for a home-cooked dinner with leftovers.”

Common pitfalls:

• Pasting everything (dilution and overflow).
• Vague goals (“analyze this”).
• No requested format (the model guesses and meanders).

One-screen recap

• Prompt engineering = designing your ask so the model can read it, keep it on the page, and answer cleanly.
• Use the 5-piece structure: Goal → Key info → Format → Constraints → Tiny example.
• Keep inputs short, labeled, and relevant.
• Show a pattern you want the model to copy.
• Ask for the form you want back. Clear beats long.

💼 401(k) Match Maximizer

Why this matters: Employer matches are guaranteed returns — yet a surprising number of people leave free money on the table. In 10 minutes, you’ll extract the exact contribution % needed to capture your full match, understand vesting, and set a reminder to keep it on track.

Step 1 — Find your plan details (2 minutes):

• Open your HR/benefits portal → Documents / Resources → download the 401(k) Plan Highlights or SPD PDF. Open the PDF and copy the Match and Contributions sections (a few paragraphs is enough).
• Or find the documents with your 401k administrator
• Or use any other resources available to you to find this document

Step 2 — Extract the match math with ChatGPT (4 minutes):

Paste the excerpt into ChatGPT (Free) with this prompt:

You are my plan explainer. From the text I paste next, extract:

1. Employer match formula (e.g., “100% on first 3%, 50% on next 2%”).
2. Maximum match cap (total % of pay employer will match).
3. Employee contribution types allowed (pre-tax, Roth, after-tax).
4. Per-paycheck % needed to get the full match (state the minimum % I must set).
5. Vesting schedule for match funds (e.g., immediate, 2-year cliff, graded).
6. Contribution limits (employee IRS annual limit; state “use IRS limit” if not in text).
7. Checklist to change my contribution in the payroll/benefits portal (2 steps).

Rules: Use only the pasted plan text; if a detail isn’t present, say “not disclosed.” Output a 1-page result with bold section headers and one final line: “Set your contribution to: X% (or higher) to capture the full employer match.”

You’ll get a plain-English one-pager with the exact % to enter.

Step 3 — Save your Match Plan & set a reminder (3 minutes):

• Paste the ChatGPT output into Google Docs (or Notion Free) and title it: 401(k) Match Plan — <Company Name>
• Open Google Calendar → create an event 2 business days before the next payroll cutoff (usually your next pay date minus 2 days):Title: Check 401(k) % — ensure at least X% for full match Description: paste the Docs link + your 2-step portal checklist.
• Add a monthly recurring reminder so you don’t drift below the match if pay or elections change.

Step 4 — (Optional) Add Roth/Pre-Tax note & vesting awareness (1 minute):

• If the plan permits Roth and pre-tax: add a one-liner in your doc about the split you prefer (e.g., 100% pre-tax or 50/50).
• Note vesting: if match is not immediate, add the vesting timeline to your doc so you know when employer dollars fully belong to you.

The Payoff

In 10 minutes, you’ll lock in free employer money every paycheck, know your vesting timeline, and have a recurring reminder so you don’t slip under the threshold. It’s the cleanest, highest-ROI optimization most people can make.

Transparency & Notes for Readers

• All tools are free: ChatGPT Free, Google Docs/Notion Free, Google Calendar.
• Accuracy: Use text directly from your plan PDF/portal; if anything is unclear, contact HR.
• Limits: Some plans have true-up rules or annualized matching — your doc will note it if present.
• Educational workflow — not financial advice.

If you follow the AI newscycle or read headlines from major news sources, you’ll notice a constant theme circulating: AI is shrinking the labor force and already starting to replace human workers. That can be a scary proposition for the future, as the seismic shift is happening all around us. That feels like a constant boulder being thrown on our shoulders as we march down an uncharted path. This is where ‘fear headlines’ can be successful at distorting our view: “AI will eliminate X million jobs.” “Entire functions automated overnight.” “Company X projects 50% less human workers in the next 18 months due to AI.” I read these headlines the same as you, but want to share an alternative perspective to help you filter through the noise as the world changes all around us. 

1) Fear is big business

There is a grim and science-backed reality baked into our 24/7 news cycle: Negativity drives online news consumption. According to a 2023 study, for a headline of average length, each additional negative word increased the click-through rate by 2.3%. Fear equals big business for media companies that rely on clicks to drive advertising revenue. Understanding this framework helps me digest AI articles without immediately developing an emotional reaction to the headline. I find it challenging to think critically when strong emotions (good or bad) are tied to a topic. Reminding myself of why a headline might be worded a certain way helps me focus on the content and formulate my own opinion.

2) What’s really impacting jobs right now

The labor market is a complicated web, with many variables that at any given time can impact the overall health of the job market. AI is definitely becoming a larger force, but by no means is AI the only reason for corporate layoffs. There are several factors that impact the broader jobs market:

• Over hiring during Covid era
• Federal policy decisions
• Interest rate environment
• Legislation
• Global macro economic conditions

AI seems to generate the most interest and becomes an easy scapegoat any time a company announces layoffs. It also seems easier for CEOs to use AI as a crutch instead of addressing some of these more complex issues (at least in the public domain) that might be impacting their business. It’s important for me to take all of these facts into play when trying to understand how real a headline is and what it could mean for me.

3) Headlines don’t always equal reality

I try to take headlines, both positive and negative, as inputs to form my own opinions and hypotheses. Specific to AI and the labor market, I run the article through a filter to try and understand if there are broader implications I should pay attention to:

• Was this AI or accounting? Sometimes it’s a pivot from overhiring, a poor decision, or a reorg that would have happened with or without AI.
• Is this an industry, company, or country trend? I try to spot trends and correlate whether or not the trend will impact the entire country, or will it be isolated to a specific industry. For example, many of the AI related layoff announcements have been from the major tech companies. We’ve yet to see this trend spread to other industries, at least not yet. 
• What is the scope and time horizon? Near-term trims can coincide with long-term reinvestment in new roles (eg., automation ops, compliance, evaluation). If cuts and hires occur together, the story is reshaping, not retreating.

This gives me a consistent framework to understand the ‘why’ behind a layoff announcement and how I can use that information in my AI journey. It’s not perfect, but has proved more helpful than scanning headlines and being influenced by fear. 

If you take away only one learning from this week, remember to focus on what you can control. The truth is, no one knows exactly how AI is going to disrupt the labor market in the short, medium or long term. All we know is that change is coming, and it’s hard to understand how significant the shift will be as we live through it in real time. It wasn’t that long ago that thousands of people across the country worked in mail rooms. Imagine what was going through their minds as email was introduced into the workplace. In a flash, an entire job function vanished from corporate America. While AI is new, disruptive, and moving at the speed of light, this is not the first time technology will reshape how companies operate. And it won’t be the last.

Goals & Milestones: