victoriagonzalezfores@lasallebuenconsejo.es
Forum Replies Created
-
Posts
-
I think the best way to learn how a machine learns, from what I observe with students in my Computer Science and Robotics classes, is to use applications like Machine Learning. The boys and girls at first think that the machine has more skills than it actually has, but as they use the application they realise that the only thing the machine actually does is handle data. Its advantage is that it is capable of using large volumes of data in a very short time, but they also realise that it is not infallible and that details as small as offering a single point of view as training, makes it incapable of differentiating between simple things.
In higher grades they learn about the bias that comes from being trained on majority data, and how the machine’s predictions are based on frequency. This can perpetuate inequalities or injustices, for example when using artificial intelligence to select personnel or make decisions where ethics may be compromised.
I think these reflections are very necessary, because they make them less manipulable people, who critically approach the use of these new technologies, but also of their data or their personal image.
I think the best way to learn how a machine learns, from what I observe with students in my Computer Science and Robotics classes, is to use applications like Machine Learning. The boys and girls at first think that the machine has more skills than it actually has, but as they use the application they realise that the only thing the machine actually does is handle data. Its advantage is that it is capable of using large volumes of data in a very short time, but they also realise that it is not infallible and that details as small as offering a single point of view as training, makes it incapable of differentiating between simple things.
In higher grades they learn about the bias that comes from being trained on majority data, and how the machine’s predictions are based on frequency. This can perpetuate inequalities or injustices, for example when using artificial intelligence to select personnel or make decisions where ethics may be compromised.
I firmly believe these reflections are very necessary, because they make them less manipulable people, who critically approach the use of these new technologies, but also of their data or their personal image.
I think the best way to learn how a machine learns, from what I observe with students in my Computer Science and Robotics classes, is to use applications like Machine Learning. The boys and girls at first think that the machine has more skills than it actually has, but as they use the application they realise that the only thing the machine actually does is handle data. Its advantage is that it is capable of using large volumes of data in a very short time, but they also realise that it is not infallible and that details as small as offering a single point of view as training, makes it incapable of differentiating between simple things.
In higher grades they learn about the bias that comes from being trained on majority data, and how the machine’s predictions are based on frequency. This can perpetuate inequalities or injustices, for example when using artificial intelligence to select personnel or make decisions where ethics may be compromised.
I think these reflections are very necessary, because they make them less manipulable people, who critically approach the use of these new technologies, but also of their data or their personal image.
I think the best way to learn how a machine learns, from what I observe with students in my Computer Science and Robotics classes, is to use applications like Machine Learning. They at first think that the machine has more skills than it actually has, but as they use the application they realise that the only thing the machine does is handle data. Its advantage is that it is capable of using large volumes of data in a very short time, but they also realise that it is not infallible and that details as small as offering a single point of view as training, makes it incapable of differentiating between simple things.
In higher grades they learn about the bias that comes from being trained on majority data, and how the machine’s predictions are based on frequency. Inequalities or injustices can be perpetuated.
I think these reflections are very necessary, because they make them less manipulable people, who critically approach the use of these new technologies, but also of their data or their personal image.