Artificial Intelligence Must Be Watched Closely: Microsoft's Brad Smith

It may sound like a joke to many, but until not long ago, the fear of Artificial Intelligence by seeing it as a threat to the human racewas a fairly widespread phobia. In fact, we already saw it in this talk in our episode one of the second season of Captcha in which we saw how it is an unfounded fear.

However, it seems that not everyone sees it that way and many think that Skynet or Matrix and the machine revolution is just around the corner. Artificial intelligence can be a threat and so at the GeekWire Summit 2019, the President of Microsoft,Brad Smith, has given his vision, somewhat pessimistic about it

Is Artificial Intelligence a threat?

Smith gave the reason for this somewhat negative view, the plane crash that occurred in the Puget Strait area, in which a plane with passengers and crew suffered an accident due to an automatic system that they could not disable from the cabin when the time came.

The software and its premise were magnificent, since it is in charge of detecting an excess in the inclination of the aircraft if it is too high, lowering the nose and decreasing the height accordingly. The problem is that in this case the measurements failed and the cabin crew could not deactivate the system with the consequent tragic end.

"

This situation has been the excuse for Brad Smith, who justifies that technological development is laudable, but must always be controlledAnd in the case of Artificial Intelligence and the systems that incorporate it, these must have a kind of panic or emergency button that allows them to be deactivated."

And for the record, it is not the only case. On a more earthly level and much closer to the vast majority of users, we have cases of people being run over by by autonomous cars.

It is about guaranteeing the primacy of human decisions at certain moments and in certain circumstances in which the decision-making of a system based on artificial intelligence is not as adequate as it should be.

Brad Smith did not cease at that point in his criticism of the use of Artificial Intelligence as we know it today. The entire society and the entire industry must become aware of the path they are taking:

"

We have seen how there are drones and war machines>It is only part of the problem that makes Smith affirm that the possible risk of an uncontrolled development of AI should be studied and automatic learning (machine learning). In fact, we already saw a similar opinion from a Google engineer."

It's a tricky problem. The development of an AI that at a given moment could not even manage the control systems themselves imposed by human beings. The crossing of ideas and theories is on the table and will give a lot to talk about.

Source | GeekWire