AI deceives people too Ke Jie is injured, but scientists say that the human brain is many times stronger

(Original title: AI Can Deceive People Too – Ke Jie Is Injured, But Scientists Say the Human Brain Is Many Times Stronger) By Jiang Xuan "I was a member of the Chinese Go National Team. In the first man-vs-machine Go match in 1996, my idol 'Chess King' Garry Kasparov defeated the computer 'Deep Blue' 4-2. When he played against Li Shishi in 2016, we thought (the computer) wouldn't win, but it did. Then Ke Jie lost this year... We felt very hurt playing Go." Recently, at the 2017 Future Forum held in Beijing, during a brain science workshop, a professional Go player admitted that he had been "hurt" by AlphaGo. In fact, his idols were also affected. On October 19, Google's AI team published a paper stating that AlphaGo's "evolutionary version" — AlphaGo Zero — learned to play Go on its own in just 40 days and defeated its predecessor, AlphaGo Master, which had previously beaten Ke Jie 3-0 in May of this year. Ke Jie sighed, "For AlphaGo’s self-improvement... humans are too much." [Image: http://i.bosscdn.com/blog/79/6b/1c/4ddb4ac932123a1853d3e7c3cc.jpeg] Last week, "robot beauty" Sophia was officially granted citizenship by Saudi Arabia — a historic moment in human history. In the near future, movies like "Silver Wings Killer 2049" explore love, hate, and the copying of human traits. For decades, artificial intelligence has advanced rapidly, causing humans to lose confidence in their own superiority. Where is the human brain now? Will people lose, and machines win? The cover of the New Yorker's October 23 issue depicted a frightening scene: robots walking side by side with humans on the street, while "small humans" kneel on the roadside, begging for them. It's a chilling image. No one can predict whether this will happen in 10 years or 100 years. At the aforementioned brain science forum, top scientists aimed to restore human dignity. They discussed research on the human brain and its relationship with artificial intelligence. Ultimately, they wanted to reassure people that there's no need to worry about computers taking over immediately. The human brain is many times stronger than any machine. [Image: http://i.bosscdn.com/blog/ca/b9/7b/6df3174aabfb29ac99cc9cfcd8.jpeg] For example, Li Feifei, chief scientist of Google Cloud AI and Machine Learning, comforted the "injured" Go players. “Go players may feel hurt, but do you think machines can feel injured? Will machines experience the same kind of pain as humans? That’s the difference between the human brain and a computer.” Interestingly, Luo Minmin, a professor at Tsinghua University’s School of Life Sciences, said, “People will lose, and machines will win. As long as a task is fixed, machines will perform better than humans.” However, she added, “Machines may eventually develop consciousness and emotions, depending on how we define them. But we’ll never find an AI that’s exactly like a human.” Stephen Hawking once warned that once AI is unleashed, it could re-design itself at an accelerating pace, making it impossible for humans to compete due to biological limitations. He believed that humans might be replaced by AI. According to Hawking’s logic, the scenario described in the New Yorker seems inevitable. However, scientists are working to prevent such a future. Luo Liqun, a Stanford professor and NAS member, explained, “The human brain has more than 10 billion neurons in the cerebral cortex. Each neuron has around 1,000 inputs and outputs. A transistor, by comparison, only has three inputs and outputs. This shows the complexity of the human brain.” Luo also mentioned recent experiments: scientists extracted signals from a specific brain region in animals, converted them into light, and injected them into another animal’s brain. The receiving animal then mimicked the movement of the source animal. This demonstrates direct brain-to-brain information exchange. Another experiment showed that stimulating a particular brain area in mice triggered aggressive behavior, showing how certain parts of the brain control complex actions. The complexity and mystery of the human brain continue to attract researchers across generations. [Image: http://i.bosscdn.com/blog/13/e7/ef/851788c797d0b8f617338c030c.jpeg] The classic sci-fi film "Silver Wings Killer 2049" explores humanity’s fear of intelligent robots. Luo Liqun explained that neuroscience studies the brain at multiple levels — from proteins and synapses to neural networks and behaviors. These integrations have led to some of the most exciting developments in the past decade. It’s worth noting that disciplines like physics, medicine, and history are closely related to brain science. The connection between computer science and neuroscience is even closer. "Deep learning" is too primitive. The breakthrough technology of "deep learning" in AI simulates the human brain's response to new stimuli. The human brain has billions of neurons that process information through connections. Similarly, AI uses artificial neural networks to simulate this process. However, scientists believe that this approach cannot fully replicate human cognition. Nicholas Turk-Browne, a psychology professor at Yale, said, “We can store a lot of data in computers, but we can’t necessarily store individual experiences.” He gave an example: if you ask me what Trump’s phone number is, I don’t know. But a computer would either say you need to create a database or search one. This highlights the difference between human memory and machine processing. Turk-Browne believes that studying human learning and memory can help improve AI systems. While AlphaGo continues to dominate in Go, discussions focus on AI's comprehension, creativity, and emotions. AI may surpass humans in computing power, but in deeper understanding, it may not match even a three-year-old child. Dr. Sophie Vandebroek from IBM Research told a reporter, “AI requires continuous learning to reach human-level understanding and reasoning. It’s still a long way to go.” [Image: http://i.bosscdn.com/blog/d2/0d/1c/a941fe3299e00cefa4ddaed4fa.jpeg] Saudi Arabia granted citizenship to the robot Sophia. Geoffrey Hinton, known as the "Godfather of Deep Learning," mentioned that deep learning has reached a bottleneck. Some scientists are exploring alternative methods, like transfer learning, which is more suitable for companies with limited data. Ding Jian, Managing Director of Jinshajiang Venture Capital, said, “Brain science is the largest unknown territory for humans and also the closest field to us. Every progress brings surprises.” He emphasized that talent is crucial for investment in cutting-edge tech, and he remains focused on the interest stage. “Deep learning is too primitive compared to the human brain,” Ding said. “The human brain is far more intelligent and magical. For now, it's unlikely that computers will surpass the human brain.”

Organic Light Emitting Diode

Organic Light Emitting Diode,Pm Oled,Oled For Medical Products,Stepper Motor Oled

ESEN HK LIMITED , https://www.esenlcd.com