The argument, in full.
Long-form analysis on enterprise AI, governance, agentic systems, technology sovereignty, and redesigning broken systems.
- Deha KuranSeptember 2026 · 8 min read
Skill Atrophy Is the Hidden Cost of AI in 2026
AI can raise output while weakening human judgment. In 2026, enterprises must treat skill atrophy as an operating risk, not a side effect.
Read article → - Deha KuranSeptember 2026 · 4 min read
Who Owns Your AI After It Goes Live?
Who maintains enterprise AI after launch? Closing the ownership gap between IT, business teams and builders requires people, funding and authority.
Read article → - Deha KuranSeptember 2026 · 3 min read
Model Distillation Explained: When Smaller AI Makes Sense
Like a chef teaching an apprentice, model distillation trains smaller AI for useful jobs. Discover when it saves resources and when ordinary software is enough.
Read article → - Deha KuranSeptember 2026 · 11 min read
AI Labs Are Slowing Down—and It’s Not Just Because of Fear of AGI
AI labs are slowing down. Beyond fear of AGI, increasingly capable agents are testing a digital world built around the limits of human attackers.
Read article → - Deha KuranSeptember 2026 · 10 min read
Forward-Deployed Engineering: Where AI Earns Its Return
Discover, validate, productize, scale: how forward-deployed AI engineering finds unmet needs, proves ROI, and builds sustainable customer value.
Read article → - Deha KuranSeptember 2026 · 6 min read
MHS Won’t Give AI Hands. FPGAs Might Give It Boundaries.
Anthropic’s MHS makes machines AI-operable, but the retrofit bottleneck is the driver ecosystem, not silicon. The FPGA’s real role is the trust boundary.
Read article → - Deha KuranAugust 2026 · 4 min read
Europe Is Not Missing Technologies. It Is Missing the Stack.
AI, quantum and photonics will not scale in Europe without energy, compute, sovereign cloud, fabrication, capital and skills.
Read article → - Deha KuranAugust 2026 · 2 min read
The Em Dash Has Somehow Become an AI Disclaimer
The em dash became an AI tell. But the deeper risk is human communication optimizing itself to the level of machine-generated content.
Read article → - Deha KuranJuly 16, 2026 · 2 min read
Who Holds the Keys to AI’s Future?
A blue door in Greece prompts questions about AI access, transparency, public understanding, and who gets to shape humanity's technological future.
Read article → - Deha KuranJune 2026 · 3 min read
AI Labs Are Becoming Enterprise Operating Layers
AI labs are moving from model providers to enterprise operating layers for agents, access, workflow, integration, and governance.
Read article → - Deha KuranMay 2026 · 4 min read
Coding Benchmarks Got Boring. ProgramBench Made Them Honest.
ProgramBench asks agents to rebuild programs from binaries and docs. The models hit the floor, and that may be the point.
Read article → - Deha KuranMay 2026 · 3 min read
World Models Are the Next Shift, Not Larger LLMs
Yann LeCun's AMI Labs just trained a world model on a single GPU. The next decade of AI may be defined by predicting reality, not the next token.
Read article → - Deha KuranMay 2026 · 3 min read
The Agent Layer Is the Hard Part: Why Most Agentic AI Pilots Don't Reach Production
43% of health systems pilot agentic AI; only ~3–4% run it in live workflows. The gap isn't a model problem — it's an Agentic OS problem.
Read article → - Deha KuranMay 2026 · 5 min read
Stop Picking AI Vendors. Start Picking Layers.
Why horizontal AI tools demand one vendor, vertical AI demands one hyperscaler, and most enterprise AI is wasted on bolt-ons that capture only 10–15%.
Read article → - Deha KuranApril 2026 · 3 min read
The Highest Tax in AI Coding: Token Micro-Economics
You're paying your agent to re-learn your codebase every morning. Why coding-agent token spend is mostly waste — and what I built to fix it.
Read article → - Deha KuranApril 2026 · 3 min read
The POC-to-Production Gap: Why Organizational Readiness Is the Real Bottleneck
AI made building a POC ridiculously fast — but productization is an organizational problem, not an R&D one. Here's why your POC graveyard keeps growing.
Read article → - Deha KuranMarch 2026 · 3 min read
The Skill Erosion Paradox: Why the Most Powerful AI Models Make Deep Skills More Valuable
As AI models grow more powerful, automation-induced skill degradation threatens the very expertise needed to oversee them.
Read article → - Deha KuranMarch 2026 · 7 min read
The AI Productivity Illusion: Why Your PoCs Are Just Moving the Problem
Why most AI initiatives are expensive ways to preserve processes that shouldn't exist — and what to do instead.
Read article → - LinkedInJanuary 24, 2025 · 6 min read
From Automation to Ingenuity: The Rise of Sapient AI
We are not moving from automation to AI. We are moving from narrow tools to systems that can surprise us. That changes everything about accountability.
Read original publication ↗ - LinkedInMay 29, 2024 · 5 min read
No More Tech Push, Please!
Tech push is what happens when engineers solve problems nobody asked them to solve. We have built an entire industry on this.
Read original publication ↗ - Philips Technology BlogMay 23, 2024 · 8 min read
Navigating the Future: Philips and the Evolving AI Regulation Landscape
The most dangerous thing about AI regulation is not that it moves too fast. It's that companies are positioning compliance as transformation.
Read original publication ↗ - LinkedInMarch 6, 2024 · 9 min read
Welcoming the Age of AI Agents
AI agents don't just automate tasks. They automate decisions. That is a fundamentally different responsibility.
Read original publication ↗