- One more Thing in AI
- Posts
- When AI Eats the Web
When AI Eats the Web
Plus: The era of mass intelligence
Renjit Philip
September 08, 2025 • Estimated Reading Time: 9 minutes
In partnership with
The best revenge is not to be like your enemy.
Date: 07-Sept-2025
Hey AI enthusiast,
And just when you think you’ve seen it all…there’s always One More Thing in AI.
In this edition:
You are going to enjoy this edition!
Best,
Renjit
PS: Founders- In case you want to prototype and test AI automation /Agentic AI deployment in your business, schedule an appointment here:
When AI Eats the Web: The Matthew Prince Playbook
When he first heard media firms cry foul about AI ripping off their content, Matthew Prince rolled his eyes. He had heard these stories during the founding years of the Web. Then the numbers landed like a brick.
What Happened
• Mathew, who is Cloudflare’s CEO, checked a decade of web data. AI-powered summaries and chatbots are cutting traffic to news sites by up to tenfold. That puts the economics of online content in jeopardy.
• With his wife, he runs a local newspaper in Park City, Utah. He saw the problem not just in charts—but in homes where journalism still matters.
• Cloudflare acts as a bodyguard for about 20 percent of the internet. In July, they blocked AI crawler bots by default. The move shields content from bots that scrape without permission.
• AI companies, even those using Cloudflare, supported the change. Prince put it bluntly: they don’t want to be the ones paying while competitors get a free ride. He saw a market-maker role opening up for Cloudflare: platforms could set crawler fees.
In Business Insider, Prince laid out three possible futures for the web:
1. Content collapse: original writing disappears.
2. Oligarchic control: few giants own what people read.
3. A cooperative model: AI firms pay creators for high-quality content, much like Spotify did for music.
• Internally, Cloudflare calls this shift “Act 4.” It follows the company’s previous chapters in security, networking, and developer tools.
Why It Matters
• It reframes how value flows online.
• It invites new models that pay creators and sustain innovation.
• It offers a roadmap for startups to build in creator-friendly systems.
Matthew Prince isn’t just sounding an alarm. He’s drafting the new deal for the internet. Startups that get in now could shape a future where creators thrive, and AI learns by paying its dues. He is in Time magazine’s list of 100 influential people in AI (2025).
How to Choose the Right Voice AI for Regulated Industries
Explore how enterprise teams are scaling Voice AI across 100+ locations—without compromising on compliance.
This guide breaks down what secure deployment really takes, from HIPAA and GDPR alignment to audit logs and real-time encryption.
See how IT, ops, and CX leaders are launching secure AI agents in weeks, not months, and reducing procurement friction with SOC 2–ready platforms.
Fidji Simo’s Next Move: Profit Meets Purpose at OpenAI
Fidji Simo joined OpenAI in May as CEO of Applications. She now leads product, engineering, and business teams. The company is growing fast, but last year it reported a $5 billion loss. Her task is to build a path to revenue while keeping OpenAI’s mission in sight. She is also in this year’s Time magazine’s list of 100 influential people in AI.
Simo has a track record of turnarounds. At Instacart, she took charge after demand dropped when the pandemic ended. Many thought the company had peaked, but she drove new growth through fresh ideas and partnerships. Before that, she spent a decade at Facebook, helping build its ad business. She turned the news feed and video tools into engines that fueled much of the company’s revenue.
OpenAI is different from most tech firms. It has a non-profit structure and board, which Simo once sat on, ensures that the company’s work in artificial intelligence serves the public good. This structure creates a constant tension between making money and protecting the mission. Simo says OpenAI will not chase engagement like social platforms. Instead, it will optimize for usefulness, building tools that help people do more.
Simo’s challenge at OpenAI is simple to state but hard to deliver. She must show the world that AI can make money and still serve society. If she gets it right, she won’t just change OpenAI’s future; she may set a model for how mission-driven tech companies can grow in the years ahead.
Key Takeaways
💡 New Leadership: OpenAI named Fidji Simo as CEO of Applications in May. She oversees product, engineering, and business.
📉 Financial Pressure: The company posted a $5 billion loss last year. Profitability is a top priority.
🚀 Proven Operator: Simo revived Instacart’s growth and built Facebook’s core ad tools.
⚖️ Mission Matters: A nonprofit board keeps OpenAI focused on serving society.
🔑 Different Playbook: Instead of chasing engagement, Simo wants AI tools that drive usefulness.
🌍 The Big Test: She must prove AI can be profitable while still serving the public good.
Did You Know? The Rise of Mass Intelligence
Over one billion people now use AI chatbots regularly. ChatGPT alone reaches over 700 million weekly users, and others like Gemini add hundreds of millions more.
A key shift is underway: AI is becoming as accessible as a Google search—ushering in what Ethan Mollick calls the era of Mass Intelligence.
Until recently, free users got access to only older, less capable AI models. The most advanced “Reasoner” models were hidden behind $20 to $200 monthly subscriptions that are often complicated to use even for experienced users.
GPT-5 changed the game. It’s not just one model, it’s a smart “router” that chooses which underlying model (from lightweight to heavyweight) to use based on task complexity. That means casual users get fast response times, while complex queries get sent to advanced “GPT-5 Thinking” engines.
AI has become dramatically cheaper and more energy-efficient. GPT-5 Nano now costs just 14 cents per million tokens, compared to ~$50 for GPT-4. And AI today consumes just 0.0003 kWh per prompt, roughly the same energy as 8–10 seconds of Netflix streaming. How much water these models use per prompt is less clear but ranges from a few drops to a fifth of a shot glass (.25mL to 5mL+), depending on the definitions of water use.
AI isn’t just more affordable; it’s easier to use. Techniques like chain-of-thought prompting are fading. AI models now often interpret your intent and go beyond the exact prompt without extra help.
Image AI also leveled up. Google’s “nano banana” (aka Gemini 2.5 Flash Image Generator) is cheap enough for free users and understands plain-language instructions, making complex image edits like “Neil Armstrong in a tuxedo walking on the moon”, surprisingly simple.
As AI spreads to billions, outcomes are wildly diverse. People form strong emotional bonds with AI, get diagnosed for diseases, write obituaries or scriptures, cheat on homework, launch businesses and sometimes face mental-health risks too.
This isn’t just a tech shift. Institution, from schools to governments, were built when intelligence was scarce.
Now, they must adapt to Mass Intelligence, learning how to harness empowered users while managing chaos, rebuilding trust, and preserving human mastery amid AI’s expansion.
References
Do you want to reach a high-quality audience?
To sponsor this Newsletter and gain customers. This newsletter is subscribed to by solopreneurs, startup founders and business leaders:
Reply