Editor’s note: This article is based on “The Fourth Revolution,” written by Luciano Floridi, a pioneer in the philosophy of information, professor of philosophy and ethical information at the University of Oxford, director of the Oxford Internet Institute, and chief consultant to Google.
Artificial intelligence research attempts to reproduce the results of human intelligent behavior and create intelligence comparable to human intelligent behavior. Artificial intelligence as a branch of engineering has achieved amazing success when interest has focused on regenerating intelligent behavior. Today, humans increasingly rely on AI-related applications (intelligent technologies) to perform tasks that might not be possible with independent or streamlined human intelligence alone. Regenerative AI will often surpass and replace human intelligence in more areas than ever before. Danish computer scientist Eziger Dikoscher’s famous remark that “the question of whether a machine can think is similar to the question of whether a submarine can swim” alluded to the shared application of regenerative artificial intelligence. The next time you’re on a plane and have a bumpy landing, remember that it could be a human pilot operating the plane, not a computer.
Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include editing robots, writing robots and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.
While the focus of interest has been on generating intelligence, however, prolific artificial intelligence as a branch of cognitive science has been disappointing. Not only did it fail to surpass human intelligence, it didn’t even join the competition. At the moment, machine intelligence is at the level of toasting, and we have no clue how to improve it. You may be slightly offended, but not surprised, when a “printer not found” warning pops up on your computer screen, even though the printer in question is sitting right next to it. The fact that Watson, an IBM supercomputer that can answer questions in natural language, beat a human opponent to win the quiz game show jeopardy in 2011 only shows that artificial intelligence can be smart, but not smart. Data miners don’t need intelligence to succeed.
The two “souls” of AI (engineering and cognitive science) often compete for intellectual superiority, academic power and financial resources. This is partly because they share a common origin and intellectual heritage: the same birth event (Dartmouth Summer SYMPOSIUM on Artificial Intelligence in 1956) and the same “father” (Alan Turing, with his computer and its computational limitations, and his famous Turing test). Simulations designed to verify that the source of the simulation has been generated, or merely matches or exceeds the behavior or performance of such intelligent sources, seem to be of little use.
The two “souls” of AI have many names, and they don’t always agree. Sometimes weak ai and strong AI, or good old AI and new/new AI, can be used to describe the difference between two “souls”. I prefer to describe it in terms of the difference between light ai and strong AI, which creates less misunderstanding. The difference in goals and outcomes has led to endless, but mostly pointless, backbiting. Defenders of AI focus on the powerful results of regenerative, engineered AI, which is the goal of weak or light AI; Ai detractors focus on productive, cognitive AI’s weak output, which is the goal of strong AI. This misunderstanding is at the root of many pointless speculations about paranormal events (artificial intelligence will one day transcend the theoretical boundaries of human intelligence).
Today, simulation and functionalism cannot be confused, because the same functions (mowing the lawn, washing the dishes, playing chess) are performed by different physical systems. Simulation and outcome are linked: mimicking intelligence will achieve the same outcome (lawn mowed, dishes washed, game won) through completely different strategies and processes. The outcome is not determined by the process. This emphasis on results is technically fascinating and successful; It is a testament to the ever-expanding use of information and communication technologies in our society. But its philosophical connotations are soporific and can be summed up as “convoluted”. Is this the end of our interest in the philosophy of artificial intelligence? I think not at all, for at least two main reasons.
Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.
First, ai has opened up vast and expansive areas of research by attempting to bypass semantic barriers and “squeeze” some information processes out of hardware and syntax. These areas challenge the conceptual rights of AI itself, but they are also interesting concepts related to the potential implications and applications of AI. Some of these innovations are called new artificial intelligence, such as installed robots, neural networks, multi-agent systems, Bayesian systems, machine learning, cellular automata, artificial living systems, and many different specialized logics. Once you get into these areas, a lot of conceptual and scientific questions don’t look the same.
The second and most important point is that in order to avoid the aforementioned divergence (engineering VS. Cognitive science, simulation vs. simulation), we must realize that AI cannot be reduced to a “natural science” or a “cultural science,” because it is an “artificial science.” This is the view of Herbert Simon, a Nobel Prize-winning economist. Artificial intelligence seeks neither a descriptive approach to the world nor a normative approach. It is dedicated to exploring the limits of building, embedding and successfully interacting with artificial intelligence in the world we live in. In other words, it aims to record the world, because such artificial intelligence is some new logical mathematical code, the new text of Galileo’s mathematical book on nature. This process of documenting the world is part of what we call building the information sphere in Chapter 2, and it is important for us to better understand how the world is changing.