You don't even need modern ML to find a set of candidate SNPs for introduction, GWAS studies of educational attainment[1] give you these SNPs with weights as pa

You don't even need modern ML to find a set of candidate SNPs for introduction, GWAS studies of educational attainment[1] give you these SNPs with weights as parameters of the statistical polygenic risk score (PRS/PGS). The problem is, intelligence is a massively polygenic trait which necessitates introduction of many SNPs (into the embryo, mind you, because most of these are likely active only during the early stages of development) to achieve life-changing effect for the person in question. Given this knowledge, the realistic way to improve human cognition would be widely available (perhaps state-sponsored) voluntary embryo selection[2] for intelligence. Embryo selection via PGS is already being applied[3][4][5] as an optional procedure in IVF, so there is no technical risk, the only impediment for now is simply social inertia, politics and lack of large-scale investment. Compared to this possibility cognitive enhancement in healthy adult individuals is a much harder problem: no solution is known as of now. Still, it would be interesting to speculate about realistic procedure of such type. It is unlikely that with sufficient effort we couldn't find some way to toggle[6] the correct genes and developmental processes in an adult, and deliver it via a gene therapy. Or maybe a cell therapy[7] could impart necessary neural progenitors into the mature tissue. Given such problems at the frontiers of the known science, it is natural to think that strong ML systems operating over large -omics datasets could offer human scientists valuable help, but the progress here looks more funding-limited than technology-limited right now. 1. https://www.nature.com/articles/nrg.2017.104 https://www.nature.com/articles/nrg.2017.104 2. https://www.gwern.net/Embryo-selection https://www.gwern.net/Embryo-selection 3. https://www.ivfbabble.com/on-the-40th-anniversary-of-the-first-ivf-in-the-usa-the-first-baby-elizabeth-jordan-carr-looks-at-how-science-today-has-produced-a-new-world-first-baby-aurea/ https://www.ivfbabble.com/on-the-40th-anniversary-of-the-fir... 4. https://www.orchidhealth.com/ https://www.orchidhealth.com/ 5. https://www.lifeview.com/ https://www.lifeview.com/ 6. https://www.frontiersin.org/articles/10.3389/fncel.2018.00460/full https://www.frontiersin.org/articles/10.3389/fncel.2018.0046... 7. https://www.pnas.org/doi/abs/10.1073/pnas.0808488105 https://www.pnas.org/doi/abs/10.1073/pnas.0808488105