360AI, a live demonstration in Zhou Hongyi, said that artificial intelligence will definitely produce consciousness!

K diagram 601360_0

  Previously releasedIt is planned to launch a demo version of ChatGPT-like technology as soon as possible.

  This judgment has brought the company’s share price up one after another. What is the big AI model that the company is developing has always been a mystery. At the strategy conference on March 29th, Zhou Hongyi made an unexpected appearance and tested the big AI model being developed.

  Zhou Hongyi said that the product is still in its infancy, and it is not yet time for the official release, so "holding the baby" in advance can satisfy everyone’s curiosity. "Neither pessimistic nor blindly optimistic. China has the ability to develop its own.Big language model, but it takes some time. "

  1. Unexpected test

  On March 29th, the Group’s strategy announcement was held in Beijing. Originally, it was planned to give a speech on the company’s security business. However, at the call of the scene, Zhou Hongyi made an unexpected appearance, tested and demonstrated the big AI model under development.

  Although the field test, but Zhou Hongyi stressed that this is not a product release, can not be public beta, can not give an account.

  360 is a hot topic in the A-share market recently.Shares. The company has made it clear on the interactive platform that the research institute of the company has been continuously investing in AIGC technologies including ChatGPT-like technologies since 2020, but so far it has only been used as a productivity tool for internal business, and the investment scale and technical level are still far behind the current ChatGPT 3, and all technical indicators can only be slightly better than ChatGPT 2. Due to the training data source and application direction, the actual effect in Chinese environment is stronger than ChatGPT 2.

  Zhou Hongyi has recently become a preacher of GPT. At the scene, he clearly believes that GPT is already the smartest person in the world. "GPT 4 is both a generalist and a specialist. Previously, AI was a proprietary task, and no AI product could really understand the world. "

  Zhou Hongyi said that a real era has begun. Before GPT, it was weak artificial intelligence, and 3.5 initiated the era of general artificial intelligence, and at the same time opened up strong artificial intelligence. In the future, it will develop into super artificial intelligence, and GPT 4 is already the prototype of super artificial intelligence.

  He believes that GPT can’t be regarded as a toy, a chat tool or a search engine. The super brain behind it represents the coming of the era of super artificial intelligence. The power of the big model has just begun.

  2. Digital immortality

  Zhou Hongyi said that with the arrival of the new industrial revolution, any industry deserves to be reshaped by artificial intelligence. "If a product does not have the blessing and change of AI, it can only be a broadsword and spear, which is likely to be hit by people’s dimensionality reduction."

  He analyzed that a lot of data is the most important asset of an enterprise and can train its own internal knowledge brain. "Internal employees rely on him for business intelligence analysis and strategic discussion. Don’t you want to have such a brain with GPT function?"

  "I used to be skeptical about digital immortality, but after ChatGPT came out, digital immortality became possible by using technologies such as big language model." Zhou Hongyi had previously talked about his change of previous cognition.

  Zhou Hongyi said on the 29th that the significance of "digital immortality" is also very important. In the past, after the death of many old experts, it was a pity that knowledge was passed down relatively slowly and lost. The digital immortality brought by GPT in the future is very important.

  Will GPT bring about mass unemployment? Zhou Hongyi thinks not.

  For the future, Zhou Hongyi said at the scene that the biggest constraint on the development of artificial intelligence is energy. "If there is a contradiction between human beings and artificial intelligence in the future, grabbing energy may be one of the important reasons."

  "Artificial intelligence produces consciousness and becomes a new species." Zhou Hongyi said that personally, he felt that GPT6-10 would be conscious for many years to come. GPT3.5 is still blind and deaf, and it needs human input. But GPT4 can already understand pictures and sounds, and will understand videos in the future. Once YouTube, Tik Tok and other videos are finished, the evolution speed will be further accelerated.

  "A few days ago, Open AI released a plug-in platform, which means that the interface has been opened, and the capability of GPT4.0 has been improved a hundred times. Even if you can’t do it yourself, you can manipulate your computer to do it, which is equivalent to having hands and feet. When he has hands and feet, eyes and ears, he has the ability of brain self-evolution. Its development speed is exponential iteration. 2023 may be the turning point of artificial intelligence. This security issue is no longer covered by security companies like 360. "

  On the 29th, Zhou Hongyi made it clear that 360 will adhere to the artificial intelligence development strategy of "Qi Fei with two wings". On the one hand, promote the development of kernel technology; On the other hand, when the finished product is semi-mature, cooperate with partners and actively seize more scenes. In terms of scenes, there is the second search engine in China, which is safe and controllable.

  He believes that the development of the big language model in 360 has three advantages. Data acquisition and cleaning, artificial knowledge training, and scene advantages, including 360 search, browser and so on.

  Thousands of experts, including Musk and Turing Prize winners, called for a suspension of super AI research and development for at least six months.

  However, on the 29th, the Future of Life Institute published an open letter calling on all AI laboratories to immediately suspend the training of AI systems more powerful than GPT-4 for at least six months.

  As of press time, this letter has been signed by 1,126 technology leaders and researchers, except Musk, Hinton andIn addition, it also includes Joshua Bengio, winner of Turing Award, and Stuart Russell, author of Artificial Intelligence: Modern Methods. Steve Wozniak, co-founder of the company, Emmad Mostak, CEO of Stability AI, and other scientific and technological leaders.