Why AI Still Forgets

Why can today's AI have a huge context window and still miss the thing that matters?

This first NeuroNewton long-form pilot separates three ideas that often get collapsed together: context, memory, and reasoning. Bigger context windows help, but they do not automatically create durable memory or faithful reasoning.

Sources:
- Liu et al., "Lost in the Middle: How Language Models Use Long Contexts" - https://arxiv.org/abs/2307.03172
- Packer et al., "MemGPT: Towards LLMs as Operating Systems" - https://arxiv.org/abs/2310.08560
- Park et al., "Generative Agents: Interactive Simulacra of Human Behavior" - https://arxiv.org/abs/2304.03442
- Yao et al., "ReAct: Synergizing Reasoning and Acting in Language Models" - https://arxiv.org/abs/2210.03629
- Turpin et al., "Language Models Don't Always Say What They Think" - https://arxiv.org/abs/2305.04388

Visual note: this video uses a generated anchor image, edited diagrams, local stock footage, and synthesized music.

Видео Why AI Still Forgets канала NeuroNewton