🔮 Exponential View #571: DeepSeek shows the future, again; drones on a learning curve; solar goes up, LLM pixels & tennis robots++

2026-04-26 · Source: Exponential View · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, extended

Summary

Ukraine has drastically accelerated its drone warfare innovation cycle to seven days, a significant reduction from the eight to nine months required for retraining soldiers returning to the front a year ago. This rapid iteration is a response to the dynamic nature of modern conflict, where adversaries quickly adapt to new technologies. Concurrently, Chinese AI labs, facing compute scarcity, are prioritizing "intelligence per token or per dollar" with models like DeepSeek V4, which is marginally less capable than GPT-5.4 but four times cheaper. This contrasts with the American AI labs' "more compute, better benchmarks" approach. The broader economic landscape also shows solar power surpassing nuclear energy in 2026, driven by technology learning curves that reduce costs, making energy a technology rather than a commodity. Automation's impact on the job market is highlighted by the travel agent industry, where routine tasks were automated, leading to a 60% job loss, but the remaining 40% shifted to higher-end, curated services, increasing their salaries.

Key takeaway

For Machine Learning Engineers and defense technologists, recognize that the "more compute, better benchmarks" paradigm is being challenged by cost-efficiency and rapid iteration. Your focus should shift towards optimizing intelligence per token or per dollar, especially given rising inference costs. Embrace agile development cycles, similar to Ukraine's seven-day drone innovation, to stay competitive and relevant in rapidly evolving technological landscapes, whether in defense or commercial AI applications.

Key insights

Rapid iteration and cost-efficiency are critical for technological advantage in both warfare and AI development.

Principles

• Technology on a learning curve gets cheaper with increased production.
• Automation shifts human value to relational, non-routine tasks.

Method

Ukraine's defense innovation involves a seven-day iteration cycle for drone warfare, continuously adapting to battlefield changes. DeepSeek's strategy focuses on maximizing real-world capability per token or per dollar, turning compute scarcity into a design specification.

In practice

• Prioritize rapid prototyping and deployment in fast-evolving fields.
• Invest in technologies with strong learning curves for long-term cost benefits.

Topics

• Drone Warfare
• Electronic Warfare
• AI Compute Efficiency
• Large Language Models
• Solar Energy Transition

Best for: Machine Learning Engineer, NLP Engineer, Entrepreneur, AI Engineer, Policy Maker, Director of AI/ML

Related on AIssential

• The Coming Drone-War Inflection in Ukraine — Autonomous drone technology is rapidly evolving on battlefields, shifting warfare from manpower to production challenges.
• 😸 You're behind on AI. Here's the recap — The AI landscape is rapidly advancing across agents, benchmarks, and hardware, driving significant shifts in market dynamics.
• AI's capability improvements haven't come from it getting less affordable — AI automation costs are not rising relative to human labor, enabling continued rapid capability expansion.
• Billions Pour into Autonomous Defense as AI Redefines Warfare — Record investments in AI-driven autonomy are fundamentally reshaping warfare, shifting capabilities towards software-defined, swarm-based operations.
• The Download: a new hunt for dark matter and Kenya’s case for going solar — Technological and scientific fields are rapidly evolving, necessitating adaptive strategies and realistic assessments of complex challenges.
• Mustafa Suleyman: AI development won’t hit a wall anytime soon—here’s why — Exponential compute growth, driven by hardware and software advances, is rapidly accelerating AI capabilities.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Exponential View.