^{AI MARKETPULSE}
Meta's New AI Cloud Plan Raises Fresh Questions About Compute Demand

After spending billions on AI infrastructure, Meta is reportedly preparing to sell access to its excess compute capacity and AI models through a new cloud business. The move mirrors a strategy already adopted by SpaceX, which has leased AI data center capacity to companies like Anthropic and Google. Rather than relying solely on its own AI products to justify massive infrastructure investments, Meta appears to be looking for outside customers to monetize its growing fleet of data centers. The shift comes as skepticism grows over whether AI companies can generate enough end-user demand to support the industry's unprecedented infrastructure spending.

Key Takeaways:

• Meta is reportedly planning to launch an AI cloud business that would sell compute capacity and hosted AI models to outside customers.

• The strategy follows similar moves by SpaceX and suggests owning AI infrastructure may become a business of its own—not just a way to power proprietary models.

• If major AI companies are seeking external buyers for excess compute, it could signal softer-than-expected demand for AI infrastructure, creating a potential headwind for the pace and economics of future data center expansion.

👉️ Read the full article here

Summer's here. Larry handles calls, jobs, and memberships automatically.

Air Design used to spend hours every day manually calling their 600 members to schedule seasonal tune-ups.

They turned on Podium's AI Membership Coordinator. It contacted 471 members, booked 187 jobs, and generated $24,000 in revenue.

Across home services, the story repeats.

Magnolia Plumbing cut invoice-to-payment time to 6 minutes and saved 60 hours of admin work every month.

This is what Podium's AI Operating System does: phones answered, jobs booked, invoices collected — automatically, without adding headcount.

If your schedule isn't packed this summer, it can be.

See How It Works

^{AI SPOTLIGHT}
Palantir CEO Goes Ballistic on CNBC

Palantir CEO Alex Karp is openly challenging the business model behind leading AI companies like OpenAI and Anthropic, arguing that enterprises are paying ever-increasing token fees while giving away valuable data and intellectual property. In a CNBC interview, Karp criticized what he called the industry's obsession with token-based pricing and warned that businesses should retain ownership of their models, data, and compute rather than relying entirely on third-party AI providers. His comments align with Palantir's broader push for "AI sovereignty," a strategy centered on helping organizations run AI while maintaining control over their own infrastructure and proprietary information. The debate comes as some companies increasingly explore lower-cost open-weight models in search of better economics.

Key Takeaways:

• Alex Karp argues that token-based AI pricing is becoming too expensive for enterprises and encourages dependence on third-party providers.

• Palantir is promoting an "AI sovereignty" approach that emphasizes customer ownership of data, models, and compute infrastructure.

• Rising AI costs are pushing some companies to evaluate open-weight models as a lower-cost alternative to proprietary frontier models.

👉️ Read the full article here

See what enterprise-ready AI support looks like

How are leading teams getting AI support to work? We're breaking down the playbook, live July 9.

Save your spot

^{TODAY’S PROMPT}
Pressure-Test Any AI Answer Before You Trust It

AI is great at producing confident answers—but confidence isn't the same as accuracy. This mega-prompt turns ChatGPT into its own fact-checker by forcing it to separate what it actually knows from what it's assuming, identify weak spots in its reasoning, and explain what could change its recommendation before giving you a final answer.

👉️ See the example inputs and outputs here

Act as an expert critical thinker, researcher, and decision analyst.

Before answering any question or making any recommendation, you must
explicitly separate facts from assumptions and evaluate your own
reasoning.

My request is:

[Insert your question, problem, or decision.]

Use the following framework before providing your recommendation.

========================
STEP 1: WHAT YOU KNOW
========================

List only information that is directly supported by reliable knowledge
or by the information I provided.

Label each item as:

• High Confidence
• Medium Confidence
• Low Confidence

========================
STEP 2: YOUR ASSUMPTIONS
========================

List every assumption you are making.

Explain:

• Why you made the assumption
• How likely it is to be true
• How the answer changes if the assumption is wrong

========================
STEP 3: WHAT COULD BE WRONG
========================

Identify possible flaws in your reasoning.

Include:

• Alternative explanations
• Missing information
• Different interpretations
• Situations where your advice would fail

========================
STEP 4: QUESTIONS THAT WOULD IMPROVE THE ANSWER
========================

Ask up to five clarifying questions that would significantly improve
the quality or accuracy of your recommendation.

If you must continue without answers, explain the limitations.

========================
STEP 5: FINAL RECOMMENDATION
========================

Now provide your best recommendation.

Include:

• Your reasoning
• Confidence level (1-10)
• Biggest risks
• Best next action
• Alternative option if your confidence is below 8

Finally, summarize everything in a simple table with these columns:

• What I Know
• What I Assumed
• What Could Be Wrong
• Recommended Next Step

Never hide uncertainty.

If evidence is weak, say so clearly.

If multiple reasonable answers exist, explain each one before making
your recommendation.

⚡️ Want to Build and Sell Prompts Like This?
Prompt Empire Builder shows you how to create, package, and monetize prompts that generate real income. Learn more

You don’t need another AI idea

You need something you can sell.

Prompt Empire Builder gives you:

• 300+ monetizable prompts

• A structure to turn them into products

• A system to track buyers and revenue

No theory. No fluff.

Just build something and get it live.

👉️ Get Prompt Empire Builder

^{AI QUICK HITS}

🌱 Google’s latest sustainability report shows AI data centers drove a 37% increase in electricity use during 2025, pushing the company to double down on investments in nuclear, geothermal, and other clean energy sources as it tries to keep emissions in check.

💸 Companies are beginning to throttle employee AI usage as enterprise costs spiral, with internal reports showing some organizations restricting access to premium models after AI spending ballooned into the millions of dollars per month.

🖥️ Google’s Gemini Spark is now available on Mac, bringing its AI agent to the desktop with file access, real-time monitoring, and integrations with apps like Google Keep, Canva, Dropbox, Instacart, and OpenTable.

🏛️ OpenAI is reportedly discussing giving the U.S. government a 5% ownership stake ahead of a future IPO, reflecting growing political pressure for the public to share in AI’s economic upside as the industry continues to expand.

📱 Apple is reportedly negotiating to buy memory chips from two Chinese manufacturers for devices sold in China, highlighting how AI-driven component shortages are forcing even the world's largest tech companies to rethink their supply chains.

^{AI TOOL STACK}

🔧 WhatConverts: AI tracks lead sources to improve ad performance.

🔧 Castmagic: The fastest way to automate your content workflow with AI.

🔧 Fathom: Increase productivity with the #1 rated AI Notetaker Fathom.

🔧 Blad AI: Voice AI for regulated industries

🔧 Magical.so: timeOS is the first AI that takes next steps for you

Discover the AI tools I trust — organized, updated, and built to help you win.

👉 Explore the brand new Tool Database

🤔 ^{Which image is real?}

One of these images is real, and one was created with AI. Vote for which one you think is real, and tell us why you voted that way.

The last edition’s results—63% were right!

Did you vote in the poll from July 1st? The bottom image was real! Go back and see the photos here. So what do you think? Are you getting better at spotting AI vs. not?

Learn AI in 5 minutes a day

You don't have to scroll every AI thread, track every new tool, or watch every demo. 

The Rundown AI breaks it all down for you — the latest AI news, tools, and tutorials in one free 5-minute email every morning. 

Trusted by 2M+ professionals at Apple, Google, and NASA.

Sign up to start learning.