>snake oil >our book Ctrl+F "scaling" gives 0 results across this essay Deep learners have their scaling laws which seem to hold across many orders of magnitu

>snake oil >our book Ctrl+F "scaling" gives 0 results across this essay Deep learners have their scaling laws which seem to hold across many orders of magnitude of model size and compute spent on training. Critics sometimes have to resort to building their argumentation along pretty esoteric tangents to avoid naming this elephant in the room. Instead of denying scaling, why shouldn't we make it work for our society's benefit? A new generation of thinkers already proposes a whole research program for life science, designed from the scaling and transfer learning fundamentals: https://markov.bio/biomedical-progress/ https://markov.bio/biomedical-progress/