 The proposed study evaluates five artificial intelligence AI techniques for predicting the electricity consumption in a single family dwelling located in the United States. The techniques considered were random forest, RF, support vector regression, SVR, extreme gradient boosting, XG boost, multi-layer perceptron, MLP, long short-term memory, LSTM, and temporal convolutional network, CONV-1D. The models were evaluated in terms of their bias and variance by using a 10-fold cross-validation technique. The results showed that the model with least dispersion was LSTM, which presented errors of minus 0.02 percent of mean bias error NMBE, and 2.76 percent of root mean squared error NRMSE, in the validation set and minus 0.54 percent of NMBE, and 4.74 percent of NRMSE in the test set. This article was authored by Moises Cordero-Castis, Daniel Vinueva, Pablo Aguilar, and others. We are article.tv, links in the description below.