I've thought about this a little bit as well, to the point of Googling 'dissolved CO2 monitor', etc. Unfortunately, I've found that for what I was thinking, the cost would be very high. My chemistry is a bit week, but couldn't you hypothetically use dissolved O2 as a proxy for [CO2]? Dissolved O2 monitors are much cheaper and more readily available than CO2 monitors. Our tanks can be though of as being representative of sea surface conditions - except that they are in equilibrium with our house ambient air - concentrations of gas in solution will depend on the solubilities of each gas, salinity, and temp. I believe [N2] and [Ar] are relatively constant in water of known temperatures, so if salinity, temp, [N2], [O2], and [Ar] are known (or can be surmised), can't [CO2] be calculated pretty easily? This is kind of a workaround, but perhaps there is potential? It seems that if the dissolved [CO2] could be detected or calculated more economically, an alk controller would be fairly straight forward.