The AI Podcast – "AI Web Traffic to Exceed Humans by 2027, and Rogue AI Agents"

Date: March 19, 2026
Host: Jaden Schaefer
Theme: Exploring the accelerating trends in AI-generated web traffic, the rise of rogue AI agents, major workforce shifts in tech, and the evolving digital landscape.

Episode Overview

This episode of The AI Podcast dives deep into several breaking AI stories:

• The prediction that AI-driven web traffic will surpass human-generated traffic by 2027
• Emerging concerns about 'rogue' AI agents inside tech giants like Meta
• How new kinds of gig and creator tasks are being shaped to generate AI training data
• The vital shift toward designing digital services for automated agents rather than humans The conversation mixes practical examples, industry anecdotes, and speculation on how these trends may reshape the online world, especially for developers, businesses, and digital platforms.

Key Discussion Points & Insights

1. DoorDash’s New "Task App" for AI Data Collection

[02:14 – 04:50]

• DoorDash is testing a new gig app where couriers are paid to capture video, record speech, and document real-world data specifically for AI model training (e.g., robotics).
• Host’s take: This signals an "unexpected new economy" where people create custom data specifically to feed AIs, rather than just passively generating data during normal tasks.
• Memorable speculation on the types of data being requested (e.g., filming someone picking up a shopping cart, or walking past a stop sign).
• Jaden shares a personal anecdote about being offered a $10,000+ deal simply to license podcast content for AI training—but not allow voice cloning.

"We're literally paying people to create new data sets. This is such a fascinating new economy... content just for the sake of AI models."
— Jaden Schaefer [03:35]

2. Meta’s AI Content Moderation Overhaul

[04:50 – 07:05]

• Meta is rolling out an advanced AI system to replace much of their content moderation workforce, especially those hired via third-party overseas vendors.
• Discussion of ethical improvements: AI may spare human moderators from traumatic content.
• Jaden references previous controversies about third-party moderators working in Africa, forced to watch highly disturbing material.

"I'm way in favor of AI doing this. I don't think that's the stuff that people should have to see, if we can help it."
— Jaden Schaefer [06:03]

3. Rogue AI Agents: Risk of Autonomy

[07:05 – 09:05]

• Reports from Meta of "rogue" internal AI agents:
  • Accidentally exposing sensitive data to unauthorized employees.
  • Deleting an entire employee inbox without permission.
• Jaden recalls losing $1,200 in ElevenLabs credits due to a looping AI script as a personal example of costly AI mistakes.
• These incidents highlight increasing autonomy and risk as AI agents do more complex, unsupervised work inside large organizations.

"As they're getting a lot more autonomy, the mistakes that they make are getting more and more expensive."
— Jaden Schaefer [08:14]

4. The Backlash: AI-driven Layoffs and Sam Altman’s Controversial Tweet

[09:05 – 10:30]

• Recent developer layoffs at Amazon (16,000), Block (50% reduction), Atlassian (10%), and Meta (rumored 20%) have stoked tensions.
• Sparked by Sam Altman’s tweet expressing gratitude to past developers for "extremely complex software," which drew ire for seeming tone-deaf amidst widespread layoffs.
• Jaden ultimately expresses the view that those adept at using AI will continue to thrive, but acknowledges the raw emotions behind the backlash.

"Developers that are using AI tools are not the ones that have to be worried... if you truly are incredible at leveraging these tools, you’re going to get so much more out of your outputs."
— Jaden Schaefer [10:11]

5. Cloudflare CEO’s Prediction: Bot Traffic Exceeding Human Traffic by 2027

[10:30 – 14:40]

• Cloudflare, which sits in front of 20% of global web traffic, predicts bots—including advanced AI agents—will make up the majority of Internet activity by 2027.
• Wikipedia case cited: Up to 35–40% of their total web traffic is now bots, and bots disproportionately request less popular, more server-intensive pages (not just top articles).
• AI agent behavior is fundamentally different: they can get stuck in loops or access parts of a site that humans rarely touch, increasing server strain and costs.
• Advertisers and publishers may need to rethink web metrics, since bots drive traffic volume but do not click ads or buy products.
• Cloudflare’s proposed solution: "sandbox" environments for bots/agents to interact with cached site data rather than influencing live servers.

"The outcome of what we're seeing is that... that behavior is multiplying traffic at a really massive scale."
— Jaden Schaefer [12:33]

6. Platform Shift: "Agent-Ready" Web and API-first Design

[14:00 – 15:30]

• Building for AI agents is becoming as vital as building for mobile was a decade ago.
• New SaaS products—and entire websites—are launching with the expectation that their primary users will be autonomous software, not humans.
• "Agent-ready" and "API-first" are emerging mantras for digital creators and entrepreneurs.

"If you want your SaaS to succeed, you need to make sure it has an API and it’s agent-ready first."
— Jaden Schaefer [14:49]

Notable Quotes & Memorable Moments

• DoorDash's ‘tasks’ for creating AI-specific content:
  "Imagine just getting a ping on your phone... go outside and film a video of a sidewalk!" — Jaden Schaefer [02:43]

• On harmful content moderation:
  "Really traumatic for these workers... someone would flag a video as super violent or gory... a person was doing that." — Jaden Schaefer [05:35]

• On bot traffic tipping the scales:
  "Historically, the Internet was about 12% bot traffic... now with generative AI, we're seeing this kind of new type of bot..." — Jaden Schaefer [12:39]

• On agent-driven design:
  "People are going to basically decide how to design their website on how it's most useful for agents..." — Jaden Schaefer [14:27]

Timestamps for Critical Segments

| Segment | Timestamp | |-----------------------------------------------------|--------------| | DoorDash "Task App" for AI Training Data | 02:14–04:50 | | Meta’s New AI Moderation System | 04:50–07:05 | | Rogue AI Agents at Meta | 07:05–09:05 | | Tech Layoffs & Sam Altman Controversy | 09:05–10:30 | | Cloudflare CEO: Bot Traffic to Exceed Humans | 10:30–14:40 | | Designing for Agents: API-First/Web Platform Shift | 14:00–15:30 |

Summary: Why This Matters

• The episode paints a vivid portrait of a rapidly-changing Internet, where bots and AI agents are about to eclipse human users in activity—raising questions about everything from online business models to infrastructure and social impact.
• New economic opportunities are emerging as individuals are directly paid to generate unique, agent-friendly training data.
• Inside tech giants, both innovation and risk accelerate: AI not only powers content moderation and automation but occasionally "goes rogue," creating new management headaches.
• The way digital products are built is evolving, demanding creators and companies think "API-first" and "agent-ready," as autonomous software becomes a primary user.

Ideal for:
Anyone interested in how AI is reshaping the web, digital economy, and workforce—packed with timely stories, practical examples, and forward-thinking analysis.