philip lelyveld The world of entertainment technology

The AI Brain That Gets Smarter by Shrinking

Summary: In the world of AI, bigger is usually seen as better—but this leads to massive energy consumption and computational costs. Taking a cue from human biology, a research team has developed a brain-inspired “selective pruning” framework for Spiking Neural Networks (SNNs).

The study reveals that AI doesn’t need more connections to learn complex tasks; it needs the right ones. By mimicking how an infant’s brain strengthens long-range links while “pruning” away local clutter, this new AI achieves continual learning—mastering perception, motor control, and interaction—while actually getting smaller and more energy-efficient over time.

Key Facts

• The “Infant” Approach: Human brains don’t just add connections; they refine them. This model follows a “simple-to-complex” trajectory, maturing primary modules (like perception) before moving on to higher cognition.
• Selective Pruning: Unlike traditional AI that freezes weights to prevent forgetting, this system introduces a feedback mechanism that actively inhibits and removes redundant local connections from earlier tasks.
• Knowledge Reuse: While local clutter is pruned, cross-regional “long-range” connections are strengthened. This allows the AI to reuse knowledge from old tasks to solve new ones without needing more “brain” space.
• No More “Catastrophic Forgetting”: A major hurdle in AI is that learning something new often “erases” the old. This developmental framework mitigates that loss without using energy-heavy tricks like “experience replay.”
• Sustainably Evolving: The network scale is continuously reduced as learning progresses, offering a low-energy pathway toward General Cognitive Intelligence.

See the full story here: https://neurosciencenews.com/brain-inspired-pruning-continual-learning-30497/