No automated systems of implanting knowledge will ever become remotely popular. As new knowledge is acquired, it must be compared against the current body for bias, validity, and even its impact on personality. This takes time. Learning is also cumulative—new knowledge builds on old—further iterating the process to a relative snail’s pace. This compounded with the native speed limitations of biological neural networks will prevent learning from any great advancement in optimization.
Valid timeframe: 50-500 years
As computing power increases, engineering with evolutionary methods (simulate an environment, randomly make changes to a design, update that design when changes move toward the desired result, repeat until satisfied) will become the most popular method of optimizing physical and mathematical models. This is an interim solution for coping with the extreme cost of intellectual resources required to design optimal algorithms that operate based on a concrete understanding of a physical model. For example, equations to minimize friction of airflow through an engine. Eventually sentient programs will use heuristic methods to generate concrete optimization algorithms at a minimal cost of time and resources.
Timeframe: 10-500 years
Technologies that owe their existence to friction will slowly disappear. Generally, friction can only be taken advantage of because of imperfections in manufacturing. Designs utilizing high friction surfaces purposefully will lose energy. Screws and nails will become obsolete. Automobiles based on friction between the ground and tires will too. Until then, the energy generated as a byproduct of uneliminable friction can be used as a limited power source.
Timeframe: 5-1000+ years
Look at massive university research projects that are currently preposterously unreasonable on any scale. These technologies will be commonplace and cheap tomorrow. Holographic storage, quantum and DNA computing, building atom-by-atom. Not because of nearly magical, unimaginable advancement, but precision refinement. Nearly all consumer electronics started with a simple concept that was refined: the cathode ray tube (some substances glow when struck with electrons), digital camera (light can be converted to electricity), optical and magnetic storage (a bunch of magnets/reflective pits can store information).
Time-independent concept