📊 Full opportunity report: The Model Is Only 10%: The Real Lesson of the New SDLC on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
A recent whitepaper from Google highlights that in AI development, the model itself accounts for only 10% of system behavior. The focus should shift to harnessing and configuring the surrounding infrastructure, which represents 90%. This insight could reshape how organizations approach AI deployment and management.
The model is only 10%
A Google whitepaper argues software’s biggest shift is from writing code to expressing intent. Its sharpest claim: the model you obsess over is the smallest part of the system — the scaffolding around it does the real work.
The clearest map yet of how serious AI development works — and mostly tool-agnostic. But it’s a Google funnel: the concepts are neutral, the on-ramps point to Gemini, Jules & the ADK. If the harness is 90% and it’s yours, your moat and your costs both live there — so own your scaffolding, route across models, and remember: AI amplifies whatever engineering culture it lands in.
Why Focusing on Harness and Configuration Changes the AI Strategy
This shift in perspective matters because it redirects organizational efforts from chasing ever-larger models to optimizing the surrounding infrastructure. By understanding that 90% of behavior is determined by how the AI system is configured and managed, companies can achieve better performance, lower costs, and improved security without constantly upgrading to new models. This approach enables more sustainable and controllable AI deployment, which is critical as AI becomes embedded in core business processes.
AI Model Risk Blueprint: Model Validation Testing | Ethical Considerations in AI Models | Integrating AI with Business Risk Plans | Real-World AI Model … Strategies | AI Governance Tools & Resource
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Reevaluating the Focus in AI Development Strategies
Historically, AI advancements have centered on developing larger, more capable models, with improvements often attributed directly to the model’s architecture. However, recent experiments and benchmarks, including those cited in the whitepaper, show that tuning the harness—prompts, tools, context, and guardrails—can outperform raw model improvements. The whitepaper situates this insight within ongoing debates about AI cost, security, and reliability, emphasizing that configuration and verification are now central to effective AI systems. This represents a paradigm shift from the ‘model-centric’ view to a ‘system-centric’ view, where the surrounding infrastructure is recognized as the primary driver of AI behavior.“The behavior you experience in AI systems is dominated by scaffolding you can build, own, and improve—it’s not just about the model.”
— Addy Osmani
YAML Made Simple: A Beginner’s Guide to Configuration and Data Structuring
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Remaining Questions About Implementation and Cost
It is not yet clear how organizations will systematically adopt this approach at scale, or how quickly the industry will shift focus from models to harness and configuration. The specific cost-benefit dynamics across different sectors and use cases remain to be validated through real-world deployment and long-term studies.
FNIRSI LCR-ST1 10KHz LCR Meter Tweezer, Mini Smart SMD Tester, ESR Meter, Multimeter, Capacitance Inductance Continuity Diode Tester, Auto Component Recognition Electronic Component Analyze
【LCR Tweezer Tester】The FNIRSI LCR-ST1 is a multifunctional and portable testing instrument that supports precise measurement of resistance,…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for AI Development and Industry Adoption
Organizations are likely to begin investing more heavily in tools, frameworks, and best practices for harnessing AI systems. Further research and case studies will clarify how configuration and verification can be optimized at scale, potentially leading to new standards and training in system engineering for AI. Monitoring industry shifts and benchmarking improvements will be key to assessing the impact of this paradigm change.
The AI Testing Framework: Quality Assurance for Machine Learning Systems
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why is the model only 10% of the system’s behavior?
According to the whitepaper, the majority of an AI system’s behavior is determined by how the model is configured, including prompts, tools, and context management, which together account for about 90%.What is ‘agentic engineering’?
It’s an approach where AI is treated as a system of components—model plus harness—focusing on configuration, verification, and context to improve performance and reliability.How does this shift affect AI development costs?
While initial setup costs for configuration and testing might be higher, long-term operational costs are lower because tuning and managing the system is cheaper and more predictable than constantly upgrading models.Will this change the way AI tools are built and sold?
Yes, it suggests that organizations should focus on building and owning their harnesses and verification tools, making these the primary source of competitive advantage rather than just relying on model providers.Source: ThorstenMeyerAI.com
