This whole will be linked to the output via some neural network.Īs for the last part, it will be the "reverse engineering" problem. Save the user inputs, along with some other information (Is this science data ? Sociology ? Politics ? Stats ? Linear algebra ?). Anonymize the data so you can't trace it back to one specific user, but save a way to know it is this user specifically. Register and save the data from a bunch of calculators around the web. We have very briefly settled that we need to make a history, but as you said : How would you do it efficiently ? And that's a very different question. How would you do that ? By having a history of, either the users specifically (an individual), or a certain cluster of people, or just everyone. A number like this doesn't help because there is an infinite number of answers, and you want to narrow your answers accordingly.Ī simple example of infinite answers would be :īut now, you need to give credits to some answers more than others. If your question is "10", it will need to know where to go, and where to search. It can't think, or whatever : but it will give you an answer related to your question. The computer is dumb, but it is very fast at it. This is not a trivial question, but it is a question nonetheless.
I do not have enough characters, so I am answering to the comment below the original post.Įxactly.
0 kommentar(er)
