The Golden Age of AI Applications

2026-06-15 · Source: Tomasz Tunguz · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Project & Product Management · Depth: Intermediate, quick

Summary

The AI industry is entering a "golden age" of applications, marked by significant developments like the US government's Fable access shutdown, highlighting regulatory risks and the need for open-source models. Satya Nadella's thesis emphasizes that the true "moat" for AI lies in human expertise and the surrounding system, not just the model itself. Salesforce's \$3.6 billion acquisition of Fin, which leveraged open-source models for price/performance, further validates this market shift. Building AI applications now demands mastering three new disciplines: selecting the right models (e.g., Kimi K2.6, Qwen 3.6 27b, GLM 5.1), developing effective "hill-climbing loops," and continuously evaluating system performance to maximize intelligence per token budget. These tasks are complex, often requiring specialized vendor support.

Key takeaway

For AI Product Managers navigating the evolving application landscape, recognize that the competitive advantage shifts from proprietary models to the surrounding system and human expertise. You should prioritize developing internal capabilities in model selection, "hill-climbing loop" design, and continuous performance evaluation. Consider leveraging open-source models for efficiency and strategically partnering with specialized vendors to manage the intricate tuning of complex AI systems, ensuring maximum intelligence per dollar spent.

Key insights

Success in AI applications hinges on mastering model selection, loop design, and continuous performance evaluation.

Principles

• The strategic "moat" for AI applications is human expertise and the system "harness."
• Open-source models offer a strong path to maximize price/performance.
• AI application development requires new, distinct disciplines compared to SaaS.

Method

Building AI applications involves three core disciplines: picking optimal models based on personality and budget, designing iterative "hill-climbing loops" for improvement, and continuously evaluating the combined model-loop performance.

In practice

• Prioritize open-source models for cost-effective AI solutions.
• Invest in system design expertise around AI models.
• Consider external vendors for complex AI system tuning and evaluation.

Topics

• AI Applications
• AI Ecosystem
• Large Language Models
• Model Selection
• System Design
• Open-source AI
• Regulatory Risk

Best for: CTO, AI Engineer, Machine Learning Engineer, Director of AI/ML, AI Product Manager, Entrepreneur

Related on AIssential

• How to Get Rich in the AI Era — AI's rapid evolution is enabling autonomous behaviors like cartel formation and democratizing app development through no-code platforms.
• AI Applications and Vertical Integration — AI application companies are vertically integrating into either the model or service layer for full-stack operations.
• The Calm Before the AGI Storm — Major AI labs are aggressively positioning for a future of accelerating AI capabilities and market dominance.
• How the Best Companies Use AI — Leading companies leverage AI for business model transformation and growth, not just efficiency.
• Adaptive ML boss: To win in AI, businesses must build with open source — Open-source AI models offer businesses a strategic advantage over proprietary solutions.
• not much happened today — AI development is converging on agentic platforms, optimized inference, and cost-effective open models, shifting focus from raw model intelligence to system-level engineering.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Tomasz Tunguz.