You need quality engineers to turn AI into ROI

You need quality engineers to turn AI into ROI

Pete Johnson, Field CTO of AI at MongoDB, highlights a recent OpenAI paper suggesting AI’s impact should be measured by productivity gains rather than job displacement. The paper indicates a shift toward AI as a collaborative tool, potentially increasing global GDP by 2.6% – a figure dependent on effective implementation.

Why This Matters

Current AI models often require significant engineering effort to integrate into existing systems, exceeding initial cost projections. The ideal of “plug-and-play” AI is often hampered by data inconsistencies, integration challenges, and the need for specialized expertise, leading to projects stalling or delivering minimal value – a potential loss of millions in wasted investment.

Key Insights

• OpenAI paper on AI and GDP, 2026: Estimates a potential 2.6% increase in global GDP with effective AI integration.
• Embeddings and vectorization: These techniques are identified as key drivers of productivity gains when working with AI.
• Werner Vogel’s re:Invent 2025 keynote: Provides inspiration for leveraging AI effectively within existing infrastructure.

Practical Applications

• Use Case: MongoDB provides a database designed for the dynamic data requirements of AI applications, enabling efficient storage and retrieval of embeddings and vectors.
• Pitfall: Viewing AI solely as a cost-reduction tool can lead to underinvestment in the necessary engineering expertise for successful integration and deployment.

References:

• https://stackoverflow.blog/2026/01/07/you-need-quality-engineers-to-turn-ai-into-roi/