23 AI Trends keeping me up at night

Key Concepts

• Vibe Coding: Using natural language and AI agent platforms (e.g., Claude Code, Google AI Studio) to build software rapidly without traditional manual coding.
• Ambient/Autonomous Businesses: Companies that operate with minimal human intervention, using AI agents to monitor markets, handle customer service, and execute tasks.
• Agent Economy: The shift from the App Store era (humans using apps) and API economy (developers connecting systems) to an era where AI agents discover, hire, and collaborate with other agents.
• Vertical AI: Specialized AI solutions targeting specific, "boring" industries (e.g., insurance, logistics, elder care) by replacing human labor costs rather than just selling software licenses.
• Outcome-Based Pricing: A shift from per-seat SaaS licensing to charging based on results delivered (e.g., per resolved ticket), enabled by AI agents performing the actual work.
• Founder-Agent Fit: The ability of a founder to orchestrate a "fleet" of AI agents effectively, acting more like a film director than a traditional manager.
• Agent Injection: A security vulnerability where malicious actors manipulate an agent’s context window or instructions to gain unauthorized access or control.

1. The "One-Hour Company" Stack

The speaker highlights a paradigm shift in company building. Previously, an MVP took months of development and significant capital. Today, using "vibe coding" tools and platforms like ideabrowser.com, a founder can validate an idea, build a product, and acquire a customer within a single morning.

• Methodology: Rapid iteration cycles where a product is built, launched, and refined by lunch.
• Key Requirement: Success in this model relies heavily on having pre-existing distribution (an email list or audience) to bypass the difficulty of customer acquisition.

2. The Rise of Ambient and Autonomous Businesses

The speaker envisions businesses that run with near-zero daily human input.

• The Arrow of Progress: Moving toward autonomous systems that handle market monitoring and execution.
• Scale: These businesses are projected to reach seven-to-eight-figure revenues with minimal human oversight.
• Ghost Teams: Future company "About" pages may feature a few humans supported by a fleet of specialized AI agents (sales, marketing, dev) that manage subtasks via serverless functions.

3. The Agent Economy (2025–2030)

The speaker argues that we are entering an era where agents hire other agents.

• Market Potential: Gartner predicts 20% of commerce will be machine-to-machine by 2030, with the agent market reaching $52 billion.
• Infrastructure Needs: There is a massive opportunity to build the "Glassdoor for AI agents"—a marketplace or reputation system to verify agent skills and reliability.
• Framework: Using tools like Paperclip (open-source), founders can create org charts where agents spin up and shut down subtasks as needed.

4. Vertical AI vs. Vertical SaaS

The speaker distinguishes between traditional SaaS and the new wave of Vertical AI.

• Vertical SaaS: Captures a fraction of IT spend via software licenses; usually results in $10M–$100M outcomes.
• Vertical AI: Taps directly into a company’s labor P&L by replacing headcount. Because it replaces human labor, the Total Addressable Market (TAM) is significantly larger (10x).
• Strategy: Focus on "boring" niches (e.g., construction, legal, accounting) that still rely on legacy processes like faxes and phone calls.

5. The Shift to Outcome-Based Pricing

The traditional "per-seat" pricing model is declining as investors and customers realize the value of software is often unclear.

• The Trend: Gartner expects 40% of enterprise SaaS to shift to outcome-based pricing by 2030.
• The Opportunity: Building businesses that charge per result (e.g., $1.50 per resolved ticket) rather than per user. The speaker suggests that converting legacy SaaS to outcome-based pricing is a billion-dollar business opportunity.

6. Scarcity and Premium Value

As AI commoditizes generic content, design, and routine analysis, value migrates toward:

• Judgment: Human-led creative direction and original, "weird" thinking that LLMs cannot replicate.
• Physicality: The "Experience Economy" (karaoke bars, live music, co-working) becomes more valuable as digital content becomes infinite.
• Human-Made Certification: Luxury brands may adopt "No AI" labels, similar to "Organic" labels in food, to signal premium status.

7. Security Risks: The Agent Attack Surface

The speaker warns that the speed of agent adoption has outpaced cybersecurity.

• Agent Injection: The new version of phishing. Instead of tricking a human, attackers poison the agent's context window or instructions.
• Digital Hygiene: Founders must perform quarterly "agent cleanses," reviewing permissions for bank accounts, emails, and data access, similar to how one manages app permissions.

Synthesis and Conclusion

The current era is described as "asymmetric," where a single founder can build a 24/7, high-margin business with minimal employees by leveraging AI agents. The speaker emphasizes that the window of opportunity is narrow—likely 12 to 24 months before competition saturates the best niches. The key takeaway is to stop waiting for stability, embrace the "new normal," and focus on building "micro-monopolies" by combining niche audiences with agent-first operations.