cool hit counter Read what Google AI has done over the years in one article_Intefrankly

Read what Google AI has done over the years in one article

The Google I/O developer conference took place as scheduled in the early hours of May 9 in Mountain View, California. At the conference, Google made several upgrades to its own products, which include the latest smartphone operating system Android P, version 1.2 of AR Core, and google maps with support for AR navigation features. Yet when the CEO entered the room and the first conversation revolved around Google's smart assistant, we already knew that the word AI would be the key word of the conference.

Artificial intelligence chip-TPU3.0 released; google smart assistant enhanced interaction and learning function; Google News introduced AI technology to deeply analyze articles and recommend them to users; self-driving and other new technologies demonstrated. The use of several AI technologies shows us Google's ambition for AI and the great technological power that Google has as a giant in the tech world.

In fact, Google has been "concerned" about AI technology for some years. Next, we'll sort out exactly where Google's extensions to AI technology have been going in recent years.

In June 2012, the Google X division connected 16,000 processors together to build a massive system with a billion nodes. The lab staff put 10 million digital photos from Youtube videos into the system to see what the amazing "Google Brain" could do. As a result, under spontaneous conditions without any instructions, the system self-learned how to recognize the face of the cat. This seemingly simple thing can be a significant step forward for the AI industry.

A billion simple is obviously insignificant for 15 billion human nodes. But this proves the possibility of artificial neural networks as well as deep learning.

In 2015, Google released an open source tool called TensorFlow. This is a second generation AI learning system based on DistBelief for development. Currently, TensorFlow has become the world's most popular open source machine learning framework, TensorFlow has features such as speed and flexibility for every developer who wants to try artificial intelligence development.

There was even a user called deepfakeapp in 2017 who used TensorFlow to build algorithms that automatically stitch face images seamlessly into videos. This algorithm has been widely used in pornographic videos, causing many people to discuss and think about morality and ethics.

In March 2016, DeepMind, a Google company, did something very interesting. They have developed an artificial intelligence program called Alpha Dog (AlphaGo). This program can deep learn and train itself on millions of expert-level Go games.

To verify the findings of the study, DeepMind brought in world Go champion Lee Sedol. Let's see if humans are good, or machines are good. The test was watched by many media and television stations around the world and the response from the public was extremely high. In the midst of five rounds of fierce competition, Alpha Dog (AlphaGo) eventually emerged victorious by a large margin of 4:1. It is enough to see that artificial intelligence has been developed at a high rate.

It was because of this incident that many people became aware of the term artificial intelligence.

In May 2016, Google unveiled its own voice assistant, the Google Smart Assistant, at the I/O developer conference. In October of the same year, Google Smart Assistant was powered on its own flagship phone - Google Pixel. Like other voice assistants, Google Smart Assistant has learning capabilities. It remembers the user's actions and provides customized content to the user. At this year's I/O conference, Google also upgraded it with improvements to user interaction. You can even use it to order a takeaway.

In June 016, Google announced that it would be introducing AI algorithms into its own search engine. According to foreign media reports, it's an algorithm called RankBrain. Google has deeply integrated this algorithm into its own engine, and the search engine can identify keywords based on what the user is searching for, thus giving the user search results that are more in line with their habits. In September of the same year, Google Translate began using the Google Neural Translation Tethering Method (GNMT). It has become more fluent and accurate in terms of fluency and meaning.

In April 2017, Google released a free drawing tool called 'AutoDraw'. Through AI, this tool can guess what the user has drawn based on the lines they have sketched, and then find the pattern from the available gallery that best matches the image in their head.

Time came to this I/O conference, where Google not only released TPU 3.0 hardware, but also introduced AI technology on apps like Google News. As of now, Google's own products are largely embedded with AI technology.

Google can have high hopes for the AI industry. In 2017, Google CEO Pichai announced that Google's strategy was changing from "mobile first" to "AI above all". From 2012 to 2017, Google has invested in and acquired more than 30 AI companies, so you can see that Google is very keen to get a monopoly in this industry of AI.

In a few short years, AI has gone from a novelty to a service for users. Google has also been working on applying AI technology to real life over the years, boldly trying to introduce AI in multiple areas to increase its development experience. It can be said that each Google application has its own set of AI system system, which is not only the result of Google's careful thinking from the user's point of view, but also a glimpse of Google's first exploration of the future direction of human life development.

I believe that with tech giants like Google leading the way, AI technology will be used more widely. Of course, we don't know what the future holds, but I believe that, after a time, AI will eventually become one with us and will serve our lives even more.

1、Jiangbei District big data to assist social governance city installed brain pain points found early
2、Li Xiaolai microblogging about blockchain sparked hot debate netizens do you understand what you are talking about yourself
3、HTC develops blockchain phone Exodus Bitcoin transactions are more secure
4、Jijis unmanned Formula One racing team will participate in the secondFSAC contest
5、The post00s are seeking jobs this years graduates

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送