Samir SawarkarTechnical publication
Blog
Technical essays, architecture notes, field reports, and durable references for AI engineering.
Jul 24, 2026 / 6 min
Why AI Gets Worse the Longer It Works
AI agents do not get tired. They get convinced—and that quiet failure mode is becoming the reliability tax of long-running AI systems.
Jul 12, 2026 / 12 min
The Last 3% Is Invisible Until You Chain It
Why you can't tell two models apart in chat — and why that gap decides whether your agent finishes its task or dies quietly at step fourteen.
Jul 11, 2026 / 12 min
Why "I Don't Know" Is Becoming an API Call
The deepest question in AI — can a model recognize its own ignorance? — is quietly being reduced to infrastructure. Here's what's shipping, what's still hard, and where practitioners should focus.
Jul 8, 2026 / 26 min
The Probabilistic Core, Deterministic Shell: A Reliability Architecture for LLM Systems
Reliability in LLM systems is not the elimination of probabilistic behavior — it is the architectural containment of it.
Jul 6, 2026 / 12 min
The 10 AI Careers Worth Betting the Next Decade On
I ranked them by automation resistance, talent scarcity, pay ceiling, and one harder question: do they survive models 100× more capable?
Jul 4, 2026 / 22 min
The OWASP LLM Top 10 Isn't a Checklist Anymore — It's Your Attack Surface Diagram
The LLM Top 10 tells you where the system can be compromised. The Agentic Top 10 tells you how that compromise propagates once the system can act.
Jul 3, 2026 / 6 min
Why Cost Per Token Is the Wrong AI Metric
Cost per token is an infrastructure metric. Cost per successful task is a business metric. Here's the one equation that connects them—and why it flips the model you should deploy.