The path to a general-purpose AI model is not to tackle the entire problem at once. A more effective strategy is to start with a highly constrained domain, like generating only Minecraft videos. Once the model works reliably in that narrow distribution, incrementally expand the training data and complexity, using each step as a foundation for the next.

A "software-only singularity," where AI recursively improves itself, is unlikely. Progress is fundamentally tied to large-scale, costly physical experiments (i.e., compute). The massive spending on experimental compute over pure researcher salaries indicates that physical experimentation, not just algorithms, remains the primary driver of breakthroughs.

Foundation models can't be trained for physics using existing literature because the data is too noisy and lacks published negative results. A physical lab is needed to generate clean data and capture the learning signal from failed experiments, which is a core thesis for Periodic Labs.

Building the first large-scale biological datasets, like the Human Cell Atlas, is a decade-long, expensive slog. However, this foundational work creates tools and knowledge that enable subsequent, larger-scale projects to be completed exponentially faster and cheaper, proving a non-linear path to discovery.

The era of guaranteed progress by simply scaling up compute and data for pre-training is ending. With massive compute now available, the bottleneck is no longer resources but fundamental ideas. The AI field is re-entering a period where novel research, not just scaling existing recipes, will drive the next breakthroughs.

The future of valuable AI lies not in models trained on the abundant public internet, but in those built on scarce, proprietary data. For fields like robotics and biology, this data doesn't exist to be scraped; it must be actively created, making the data generation process itself the key competitive moat.

To make genuine scientific breakthroughs, an AI needs to learn the abstract reasoning strategies and mental models of expert scientists. This involves teaching it higher-level concepts, such as thinking in terms of symmetries, a core principle in physics that current models lack.

The most fundamental challenge in AI today is not scale or architecture, but the fact that models generalize dramatically worse than humans. Solving this sample efficiency and robustness problem is the true key to unlocking the next level of AI capabilities and real-world impact.

Current LLMs fail at science because they lack the ability to iterate. True scientific inquiry is a loop: form a hypothesis, conduct an experiment, analyze the result (even if incorrect), and refine. AI needs this same iterative capability with the real world to make genuine discoveries.