Correcting ecosystems in ponds and in global societies.

Alien Professor Asks Students to Create water filtration system. Human cleans Entire ocean in a Day at https://www.youtube.com/watch?v=Yd3xhkzWisQ

He worries what others think, (he wants a nod from a coworker) and seeks their approval. That biases his actions so he experiments in secret and does not have others check his work ahead of time. A test in one part of a lake is not likely to be representative. If it were as polluted with many different compounds, the whole ecosystem is likely full of holes where critical species lived as part of a whole. So fixing just one is not possible without correcting and nurturing all related species. If those have been made extinct it means designing and simulating new species to fill in parts of what is now missing. The whole is supposed to work for hundreds or thousand or millions of years. It cannot be simulated in a few hours or days – and certainly not without longer term data on interactions and measurement of actual distribution of species and resources.

The whole notion of having students compete so one is patted on the back and the rest are “losers” is a huge mistake. I do not want to take your motivation out of your story, but simply asking them to work in a team to solve it, actually talk to the people who are near the lake and will be affected from then on. It is polluted then some agency some company is doing that and needs to be stopped or corrected. Treat the people in the company and that whole company as a new sub problem, it can be solved as part of the whole.

The simulations are too simplistic. Even in a story you can simply assign 1000 AIs each to a separate part of the whole and have the team monitor and randomly audit parts, if the humans and aliens are able to add value to AIs simulations and calibrations and verification tasks.

I see groups every day working on this and larger problems. It is not their individual brilliance and “winning” that matters, but their willingness to look at the whole. That seems to depend on letting the AIs be responsible for checking all the parts not letting a few humans or aliens make all the choices of methods and tests.

Computer time for independent AIs is cheap compared to utter loss and destruction of lake sized or global ecosystem — including whole countries or planets of humans or aliens.