The CEO of the company spoke about Google's transition to the era of Artificial Intelligence, Sundar Pichai, opening Google's developer conference, “I/O 2017”.
In his speech, he said, "I'm in Google for 13 years and it's remarkable how its founding mission, making information accessible and useful, is just as timely. From the beginning, we have applied computer science and its conclusions to solving complex problems, even when the technology around us has caused dramatic changes.
Τα πιο περίπλοκα προβλήματα τείνουν να είναι αυτά που επηρεάζουν την καθημερινότητα των ανθρώπων και είναι συναρπαστικό να βλέπουμε πόσοι άνθρωποι έχουν κάνει την Google μέρος της καθημερινότητας τους: Υπάρχουν 2 δισεκατομμύρια ενεργές Appliances Android, το YouTube δεν έχει μόνο ένα δισεκατομμύριο χρήστες αλλά και ένα δισεκατομμύριο ώρες παρακολούθησης κάθε μέρα και οι χάρτες της Google χρησιμοποιούνται για πλοήγηση σε απόσταση 1 δισεκατομμυρίων χιλιομέτρων κάθε μέρα. Αυτή η ανάπτυξη θα ήταν αδύνατη χωρίς την πληροφορική μετάβαση στα κινητά, κάτι που έκανε την Google να επανασχεδιάσει όλα τα προϊόντά της- επανεφεύροντάς τα έτσι ώστε να αντικατοπτρίζουν τα νέα μοντέλα διάδρασης, όπως οι οθόνες πολλαπλής αφής.
We are now witnessing a new transition in computing: the shift from the dominance of mobile to that of artificial intelligence. As in the past, this change forces Google to reinvent its products in an effort to create one world που επιτρέπει μία πιο φυσική και απρόσκοπτη αλληλεπίδραση με την τεχνολογία. Σκεφτείτε την αναζήτηση Google: αναπτύχθηκε πάνω στη δυνατότητα να αναγνωρίζει κείμενο σε σελίδες του ίντερνετ. Αλλά πλέον, μέσα από την εφαρμογή της μηχανικής μάθησης, μπορούμε να κάνουμε τις εικόνες, τις φωτογραφίες και τα video χρήσιμα στους χρήστες όσο ποτέ άλλοτε. Η κάμερα του κινητού σας τηλεφώνου μπορεί να “δει”, μπορείτε να μιλήσετε στο κινητό σας και αυτό να απαντήσει – η ομιλία και η όραση γίνονται τόσο σημαντικές στην πληροφορική όσο και το keyboard και οι οθόνες πολλαπλής αφής.
Ο Digital assistant of Google is a tangible example of these developments. It is already availables on 100 million devices, and getting more useful every day. The Google Home can now distinguish different voices, giving users a personalized experience each time they interact with the device. Google is also now able to turn the mobile camera into a tool for handling everyday tasks.
The Google Lens is a series of computing capabilities that, by understanding what you see, can help you take your next steps. For example, if you are on the floor of a friend's apartment to see a large and complex Wi-Fi code behind a router, your mobile phone can now recognize the code, understand that you are trying to connect to a Wi-Fi network -FI and connect you automatically. The key point is that you do not need to know anything to do it. Interaction and experience becomes much simpler than the process of copying and pasting passwords from one application to another on a smartphone. New Google Lens features will be added to Digital Assistant and Google Photos and then you can expect them to become available in other products.
All of this requires the appropriate IT infrastructure. Last year at I / O, Google announced the creation of the first generation of dedicated processors TPU (Tensor Processing Unit), which allows machine learning algorithms to run faster and more efficiently. Today, we announce the new generation of TPU processors: their Cloud TPUs, που βελτιστοποιούν τις λειτουργίες εξαγωγής συμπερασμάτων, τη διαδικασία της μηχανικής εκπαίδευσης και μπορούν να επεξεργαστούν πολύ περισσότερες information. Η Google φέρνει τους επεξεργαστές Cloud TPUs, στην platform and Google Compute Engine so companies and developers can take advantage of it.
It is important for us to make all these developments beneficial to everyone and not just Google product users. Great achievements in complex social problems can be realized if scientists and engineers have better, more powerful information tools for their research. But today, there are many obstacles to making all of this a reality.
This is the motive behind it Artificial Intelligence of Google, merging all of Google's Artificial Intelligence initiatives into one, which can overcome these barriers and accelerate the efforts of researchers, developers and companies.
One way we hope to make artificial intelligence more accessible is to simplify the creation of machine learning models, also called neural networks. Today, the design of neural networks is very time consuming and requires such cognitive expertise that it limits their use to a small community of computer scientists and engineers. For this reason, Google has created an application called AutoML and which proves that it is possible for neural networks to design new neural networks. We at Google hope that AutoML will get the feature planning of neural networks that today only PhD holders in the relevant subject have and in 3 to 5 years to offer the possibility to hundreds of thousands of programmers to design neural networks based on their needs.
In addition, Google's Artificial Intelligence researchers (Google.ai) work with scientists from many academic disciplines and developers to find solutions to a number of problems, having so far delivered very satisfactory results. We used mechanical learning to improve the algorithm that detects it spread of breast cancer to adjacent lymph nodes. We also found that Artificial Intelligence minimizes time and maximizes the accuracy with which researchers can explore properties of molecules but also the ones human genomes.
This transition to the new world of Artificial Intelligence is not limited to creating futuristic devices or conducting a frontier research. We believe that millions of people can help today through the full democratization of access to information and the emergence of new opportunities.
For example, almost half of employers in the United States say they have trouble filling vacancies. At the same time, the unemployed often do not know about the opening of a position almost next to them, because the nature of the ads - frequent changes, low traffic, inaccuracies in job titles - were the reason why search engines do not recognize them. Through a new initiative, Find Jobs (Google for Jobs), we hope to link companies with potential workers and help job-seekers find new opportunities. As part of this effort, it will launch a new search function that will help people look for jobs based on their experience and wage levels. Jobs include sectors where it has traditionally been much more difficult to look for and classify jobs, such as service and retail.
It is a source of inspiration for us to envision the fruits Artificial Intelligence can deliver. There is still a long way to go before we really get into the world of Artificial Intelligence, but the more we can work to provide greater access to technology - offering tools to people - so soon all will benefit. "