Can the machines think? Difficult question, once the great question of a humanity that received the “Computational Machinery and Intelligence” of Alan Turing to start the revolution and bring us a digital era in which, today, virtually all of us would answer affirmatively to that question.
Artificial intelligence is now part of our daily life, even more since smartphones have been proposed to use it for a host of daily tasks such as taking the best picture, optimizing their energy performance or helping us to perform tasks in a simple way just by speaking to a assistant.
The phones want to “think” as humans to anticipate our needs, understand the space in applications of augmented reality or virtual, recognize objects or help us solve problems before we even have them, and the AI is putting much in its development to understand us to us and our brains.
And even with all of this, however, it would be difficult to define if a machine can think although it apparently does, because exhibiting that intelligent behavior similar to human or preventing us from distinguishing human machines is precisely what a Turing test tries to evaluate. No artificial intelligence could break until the year 2014.
‘The Imitation Game’, or how to distinguish a machine from a human being
Much development has lived the AI since the time of Turing, but the mathematical genius that changed the course of a great war has had enough to do in slow but unstoppable evolution of the AI. In fact, that skill test for the machines that he himself called ‘The Imitation Game’ has always been taken as an irrefutable demonstration that a machine is capable of thinking and acting like a human being.
It may seem like a lie in six decades of research and development, but it is true that the world had to wait until 2014 to see an AI approving the Turing test, and it was the Eugene Goostman bot that could fool 30 Thursdays and 5 machines in a conversational game where he imitated the personality of a Ukrainian teenager.
And yet, in the same test done at the University of Reading it was shown that in 66.7% of the cases the machines do not manage to deceive the judges, specifying the own lack of knowledge of a teenager to pass the Turing test. It seems clear then that there will not be a revolution of the machines soon, and not in vain many experts still doubt that these tests can show whether a machine thinks or does not think for itself.
Surely in the cognitive intelligence in which Huawei works, and in that ability to obtain knowledge and use them to solve problems correctly, is the key to a machine capable of thinking, and this is where the Chinese manufacturer takes advantage thanks to a chipset HiSilicon Kirin 980, which with its double NPU and its manufacturing technology of 7 nanometers is placed at the forefront in the advances of the AI.
Nobody doubts that in the medium term we will see devices capable of retaining knowledge from our way of using them and what we need from them, to then apply what they have learned in offering us functionality anticipating us.
In a few years autonomous cars will be common and the AI will be within our mobile devices to do almost everything, but these improvements are not an uncertain future but the present, as Huawei already has the Mate 20 of those who probably they are the most advanced telephones in the world in artificial intelligence.
Greater power for activities related to machine learning means progressing much faster and doing it better managing resources, scarce in a battery-operated device, so Huawei’s latest top-of-the-range phones with their double NPU are placed in the pole position of the AI and the developments related to it.
Not in vain, the Huawei Mate 20 are able to analyze what your camera sees to understand the scene that the user wants to photograph, and thus adapt between more than 1,500 scenarios and configurations to obtain the best result without direct intervention of the user.
This time it can be affirmed, and that is that the intelligent camera mode of the Huawei Mate 20 is really intelligent, being able to help us with the focus, with the lighting, with the framing, with the white balance, with the saturation, etc. …