AI is revolutionizing protein structure prediction and drug discovery through rapid predictions and automated workflows, but experimental validation and regulatory/methodological hurdles still require careful handling.

The strongest AI impact appears where outcomes are checkable or automatable, while transformative changes in wet-lab domains may unfold more slowly.

Software engineering is undergoing clear AI-driven disruption as developers increasingly guide and review AI-written code, with studies showing notable automation gains in coding tasks.

AI-generated mathematical proofs and publications are arriving more quickly, spurring debates about validity, provenance, and the risk of unsustainable publication rates.

AI-powered modeling is becoming central across chemistry, materials science, genomics, neuroscience, weather forecasting, and instrument data analysis, often delivering faster initial results and energy efficiency.

Regulatory writing and clinical trial documentation are increasingly automated, potentially cutting drafting time, though concerns about accuracy and oversight remain.

AI is accelerating progress across many scientific domains, but effects vary by task and are challenging to measure at scale.

Wet-lab experimentation remains slower to automate due to tacit knowledge, experimental nuance, and the essential role of human judgment in decision-making.