Statistics at a scale that stopped behaving like statistics.
Geoffrey Hinton and colleagues established core methods—backpropagation for representation learning, and Boltzmann machines for unsupervised deep generative modelling—between 1985 and 1995. Their work was grounded in statistical mechanics, experimentally verified on small-scale networks, and explicitly recognised as foundational for machine learning with artificial neural networks. It did not deliver production AI, nor claim to.
Bacteria can pick up genes from other cells by transformation, transduction or conjugation, not only by inheriting them. Lederberg and Zinder showed in 1951 that a virus could carry genes from one strain of Salmonella to another, and Lederberg's earlier work with Edward Tatum showed that E. coli can exchange genes by conjugation. The material treats the process as the main route by which antibiotic resistance spreads among bacteria.
Doudna and Charpentier’s 2012 work established CRISPR-Cas9 as the first programmable, RNA-guided DNA cleavage system. It relies on Cas9 protein and a synthetic guide RNA chimera. It works in vitro. It does not demonstrate editing in living organisms or repair outcomes. It is the origin of all subsequent CRISPR applications — not a refinement, but the first functional design.
3D printing is a layer-by-layer fabrication method rooted in Chuck Hull’s 1984 stereolithography patent. It evolved from rapid prototyping in the 1980s into an industrial production option by 2019. Its real advantage is enabling previously impossible geometries—like hollow or truss-structured parts—that cut weight and waste. Fused deposition modeling (FDM) was the most common process as of 2020. It does not replace conventional manufacturing at scale. It does not solve material consistency, certification, or throughput problems. Its impact is precise, not pervasive.