20 fundamental questions shaping AI's future

Place your bets...

Sifting through the torrents of AI coverage for the newsletter each week, I’m finding that it’s getting exponentially harder to keep my own OODA loop functioning to maintain situational awareness on what’s going on at a macro level.

So here’s a subscriber-only post pulling out my current thinking on the main axes of uncertainty which capture the direction of travel for AI in mid-2025.

“Vast solar-powered data centers stretching endlessly toward the horizon, massive rectangular buildings with gleaming solar panel roofs reflecting sunlight, geometric patterns of server facilities disappearing into the distance, majestic mountain range silhouetted against a brilliant blue sky, endless rows of infrastructure creating dramatic perspective lines, golden sunlight casting long shadows, epic scale showing the immense energy requirements of AI, photorealistic architectural photography style, wide-angle lens perspective, crisp details in foreground fading to atmospheric haze at horizon” — Midjourney 7

While the headlines focus on the latest model releases and funding rounds, the real story lies in a series of fundamental tensions that will shape the next few years of AI development. While the underlying AI phenomenon is technological, these aren't just technical questions—they're strategic, economic, and philosophical divides that will determine who “wins”, who “loses”, and what kind of future is being built around us.

Here are 20 critical dimensions where the industry is placing its bets:

• Technical foundations

  • The Scaling Hypothesis

  • GPU Supremacy

  • Data Sufficiency

  • Reasoning Emergence

• Infrastructure wars

  • Energy Capacity

  • Concentrated vs. Distributed Intelligence

  • Inference vs. Training Compute

• Capability ceilings

  • Intelligent Agents

  • Intelligent Autonomous Robots

  • Multimodal Integration

• Economic battleground

  • Economic Acceleration

  • Global AI Oligopoly

  • Investment Bubble

• Safety and control

  • AI Safety/Alignment

  • Regulatory Capture

  • Effective International Regulation

• The human factor

  • Human-AI Collaboration

  • Democratic Distribution

  • Democratic Control and Direction

• AI and Earth's biosphere

  • Symbiotic AI

Let’s outline the key axis endpoints for each of these dimensions below…