Turn AI into your "life mentor"? Google is revealed to be "plotting" dozens of shocking updates

Text: Juny Editor: VickyXiao

Image source: Generated by Unbounded AI tool

If it weren't for Microsoft and OpenAI's pressing step by step, Google should still be walking leisurely in the debugging of artificial intelligence technology.

For a long time, Google has played a leading role in the technical research of artificial intelligence, but it has always held a responsible and cautious attitude towards the large-scale application of artificial intelligence technology. However, the sudden pressure of market competition seems to have pushed Google into a corner a little bit.

In April this year, Google announced the merger of DeepMind and Google Brain, an artificial intelligence laboratory. generative AI tools. **Including personal tools that provide users with life advice, planning and psychological counseling, as well as productivity tools for professional scenarios, etc.

Among them, the functions related to personal emotional guidance and psychological counseling are particularly noticeable, because this means that Google seems to have begun to let go of its hands and feet to more sensitive areas, trying to use more aggressive AI strategies to snatch more users.

Picture taken from The New York Times

**In the function test of AI personal "tutor", "She" is going to be reflected in reality? **

**According to the disclosure, one of the key functions that Google is currently testing is to allow artificial intelligence to provide users with personalized suggestions based on their personal circumstances (including health, emotions, finances, etc.). **For example, according to your exercise and weight loss goals, it can help you make exercise or diet plans, help you create financial budgets based on tasks, etc., and can also help you deal with interpersonal conflicts and give emotional guidance.

Google has signed a contract with Scale AI, which focuses on training and validating artificial intelligence software, to specifically test these tools and verify the safety and rationality of these tools to provide privacy-related recommendations. More than 100 doctoral researchers have participated in the project among.

**Different from current chatbots such as ChatGPT and Bard, these new features will provide more "humanized" recommendations for individual users. **For example, when you are unable to attend your friend's island wedding because of your lack of money, it will help you analyze the situation and propose solutions, and it can also accompany and answer various relationship problems.

This kind of function reminds people of a science fiction movie "Her" from ten years ago. The movie tells the story of a letter writer who is in a marital crisis, tries to use artificial intelligence to help him relieve psychological pressure and provide counseling, and finally falls in love with artificial intelligence.

The picture is from the screenshot of the movie "Her", and the copyright belongs to the original author

With the development of technology, in addition to using artificial intelligence as an auxiliary tool, more and more people hope to find "resonance" in the virtual world, find care and love, and further develop AI assistants into AI "friends". **However, due to the large number of risks related to personal privacy and emotional intelligence, this field has not been vigorously developed at present. **

It is understood that in a high-level internal meeting of Google in December last year, experts from Google's artificial intelligence security department also specifically emphasized the dangers of people being too emotional about chatbots. When Bard officially launched in March, it was also limited to providing users with medical, financial or legal advice.

But for now, Google is still trying to discover the stronger capabilities of artificial intelligence internally. However, there is no clear answer on whether these tools will be released to the public soon. A spokesperson for Google DeepMind responded: "Google continues to conduct security assessments of various products and functions, but the assessment does not represent our product route. picture."

Accelerating the pace of AI applications, Google DeepMind began to make all-round efforts

In addition to the AI "emotional mentor" function, this document also shows that Google Brain and DeepMind are also evaluating and debugging many AI tools that can be used in work scenarios, including content-based AI tools for scientific fields and creative writing. And analysis tools for data analysis, text recognition, etc., covering various industries.

In recent months, **Google has significantly accelerated the pace of launching various AI functions for specific scenarios. **

For example, last month Google launched a product that uses AI to produce news reports, and has already demonstrated it to executives of well-known news corporations such as the New York Times and the Washington Post. The tool, tentatively called Genesis inside Google, can write news stories and perform article editing based on real-time news information.

Yesterday, Google made another series of updates to the Search Generation Experience (SGE). For example, in order to help users understand a complex concept faster and better, when they hover over some words in AI search responses, they can see related definitions, pictures or diagrams.

The picture comes from Google official

For the field of code development, SGE now also supports the generation of more complex code fragments, including functions, classes, and loops. In addition, in order to make it easier for people to obtain key information when browsing the web, the SGE while browsing function is launched, allowing users to click to view the list of article points automatically generated by AI when visiting a web page, and provide a direct jump option .

The picture comes from Google official

At present, it is generally believed that **, after the merger of DeepMind and Google Brain, will further promote the emergence of Google's professional AI tools. **

Before the merger, the research directions of Google's two major AI research departments actually had their own emphasis. DeepMind's research is relatively independent, and its achievements are mainly applied in specific scientific research fields, from AlphaGo, the earliest Go robot, to AlphaZero, a general-purpose game system that has learned almost all two-player games, to being used by millions of biologists and researchers today. and AlphaFold, a protein folding simulation model used by medical researchers, et al. Google Brain, on the other hand, focuses more on the development of generative AI toolsets and large models, with the main goal of integrating AI technology into Google's products.

After the combination of the two, technical research and market application will be more deeply combined.

Demis Hassabis, CEO of Google DeepMind, said in a recent interview that since AI technology has reached the stage where it can really go out of the research laboratory, DeepMind also needs to change its research methods and increase its focus on the product level. Therefore, the merger of the two departments is the general trend and will accelerate the emergence of next-generation AI products and experiences.

In addition, Demis also said that the type of AI is much more than generative AI. **While generative AI seems to be the hottest thing right now, other capabilities of AI such as planning, deep and reinforcement learning, reasoning, etc. will return again in the next wave, together forming a more powerful system. **At that time, social innovation driven by AI will fully arrive.

*References:

Note: The cover image is from Indiatoday, and the copyright belongs to the original author. If you do not agree to use, please contact us as soon as possible, we will delete it immediately.