The assistant is a powerful example of these advances at work. It’s already on 100 million devices and becomes more useful every day. We can now distinguish between different voices in Google Home, which allows people to have a more personalized experience when they interact with the device. We are now also able to make the smartphone camera a tool to get things done. Google Lens is a set of vision -based IT capacities that can understand what you are looking at and help you act according to this information. If you have crawled on the floor of a friend’s apartment to see a long, complicated Wi-Fi password on the back of a router, your phone can now recognize the password, see that you are trying to log into a Wi-Fi network and connect automatically. We will first bring Google Lens capabilities to the assistant and Google Photos and you can expect it to also go to other products.
(Warning, to come and come to come !!!)
All this requires the right IT architecture. Last year at E / S, we announced the first generation of our TPUs, which allow us to execute our automatic learning algorithms faster and more efficiently. Today, we have announced our next generation of TPUS – COUD TPU, which are optimized both for inference and training and can process a lot of information. We will bring Cloud TPUs to the Google Compute engine so that companies and developers can take advantage of them.
It is important for us to better work these advances for everyone – not just for users of Google Products. We believe that enormous breakthroughs in complex social problems will be possible if scientists and engineers can have computer tools and more powerful and more powerful research at hand. But today, there are too many obstacles to what it happens.
It is the motivation behind Google.ai, which attracts all our AI initiatives into a single effort that can reduce these barriers and accelerate the way researchers, developers and businesses work in this area.
One way we hope to make AI more accessible is to simplify the creation of automatic learning models called neural networks. Today, the design of neural nets is extremely demanding in time and requires expertise that limits its use to a smaller community of scientists and engineers. This is why we have created an approach called autuml, showing that it is possible for neural networks to design neuronal nets. We hope that Automl will have a capacity that some doctoral students will have today and will allow in three to five years for hundreds of thousands of developers to design new neuronal nets for their particular needs.
In addition, Google.ai has associated Google researchers with scientists and developers to solve problems in a range of disciplines, with promising results. We used ML to improve the algorithm that detects the spread of breast cancer to adjacent lymph nodes. We have also seen AI make progress over time and the precision with which researchers can guess the properties of molecules and even sequence on human genome.
This change is not only to build futuristic devices or carry out advanced research. We also believe that this can help millions of people today by democratizing access to information and upgrading new opportunities. For example, almost half of us, employers, say they always have problems to fill out open positions. Meanwhile, job seekers often do not know that there is a job that opens right at the corner of the street, because the nature of jobs – high turnover, low traffic, inconsistency in employment titles – made them difficult to classify search engines. Thanks to a new initiative, Google for Jobs, we hope to connect companies with potential employees and help job seekers find new opportunities. As part of this effort, we will launch new feature in research in the coming weeks that helps people seek jobs through experience and wages, including jobs that have traditionally been much more difficult to search and classify, such as retail services.
It is inspiring to see how AI starts to bear fruit that people can really taste. There is still a long way to go before we really are a world first, but the more we can work to democratize access to technology – both in terms of tools that people can use and the way we apply – earlier everyone will benefit.
To find out more about the many other ads on Google I / S – for Android, and the photos, and VR, and more, please consult our latest stories.