What Google really does: take 2

    Preface: my previous post on this topic fell into the scrap - on the hub, alas, there is no longer a “link” section, and even a partial copy-paste in the form of a seed with a link to the original text is prohibited by the rules. Nevertheless, I consider the incident to be very important, therefore I risk once again submitting this text for discussion. To stay within the rules, I will try to retell the main text in my own words.

    So, from the middle of December last year, the famous inventor and futurologist Ray Kurzweil began working on Google. As Oleg Paramonov writes in his article (and this is the main text that I wanted to refer to), this event is out of the ordinary and, I’m not afraid of this word, is revolutionary.


    Let's first understand why Google generally artificial intelligence. Amid news about Android, hydroelectric power plants, etc., etc., it is easy to forget that Google is primarily a search. Actually, it is he who brings the lion's share of income. But what is tomorrow's search?

    Quote 1:
    “Back in 2000, Larry Page, one of the founders of Google, announced that artificial intelligence would be the ideal version of the search engine.
    Quote 2 (from a 2006 Google internal document):
    “To be the best in search, we must create a world-class artificial intelligence research center.”

    Why artificial intelligence? The fact is that the existing search does not understand the essence of the query. He simply searches by keywords, generating results based on the reputation of the page. For the search revolution, Google must understand what exactly you are looking for. At the same time, shoveling huge arrays of data, consisting not only of the page index (here it is worth remembering also the Deep Web).

    According to Oleg, if you look at the activities of Google from this point of view, you will understand the reasons for the appearance of some non-profitable Google projects. Some typical examples are given:

    1. Almost 10 years ago, Google launched a large-scale project to crawl library books. Quarrel with copyright holders but did not stop. As science historian George Dyson writes, in 2005 he was invited to give lectures for Google employees, and one of them said that, again, the quote: “The digitization of books was not started so that people read them:“ We scan books so that to be read by artificial intelligence. "
    2. Street View project. Most of the data collected by company cars is not accessible to users. But Google’s self-driving car looks at the world with almost the same sensors
    3. The GOOG-411 project, a voice guide receiving requests and giving answers in a voice form. The service is unprofitable, but its task in another is to accumulate a database of voice queries for further training of a more advanced system.

    But even with the foregoing, the employment of Kurt Kurzweil is an extraordinary event. The fact is that, again, the quote: “some consider him a brilliant seer, others - almost the founder of the sect.”

    The person who was the first in the world to create a text recognition system, and the creator of one of the first synthesizers capable of accurately simulating the sound of traditional musical instruments, is strongly associated with the term “singularity”. What it is. It is assumed that “when the limitations of the human brain are overcome, technological progress will go into a tailspin. A chain reaction will begin: powerful artificial intelligence will quickly figure out how to build an even more powerful system, it will continue to improve - and so on to infinity (or, rather, until the physical limit is reached). After that, the changes will become so quick that even the most approximate forecasts of what will happen next lose their meaning. This point of history is proposed to be called a “technological singularity”.

    According to Kurzweil, by 2030 (according to Moore’s most famous law) computers will become so complex that they can imitate the functioning of the human brain. This is where the singularity comes. But Google is certainly not interested in this. The search engine is interested in developments in the field of artificial intelligence. It’s important for us to understand where we are going. And since Kurzweil considers the appearance of a singularity possible, then we will move in that direction.

    Also popular now: