The Prospect of Generality in Artificial Intelligence

Artificial intelligence is a term that is both widely used and loosely defined. Artificial intelligence seeks to make computers do the sorts of things that minds can do, which involves diverse information-processing capacities, psychological and biological skills and the effect of AI can be found nearly everywhere.

Nowadays, the advent of mobile device and wireless internet has magnified the advantages of artificial intelligence more with networked computers becoming commonplace. Like human brains, AI acts as one large intelligent brain with much distribution, and each part deals with one specific task and lead to distributed intelligence. Connectivity is an enormous advantage for artificial intelligence over human intelligence. And specialism is spread and information can flow freely and rapidly (Alpaydin, 2016). Besides, databases and digital media have taken the place of printing on paper as the main medium of information storage. Now, it is the data which defines what to do next. AI helps human beings to extract certain rules from data. Since human behaviors are not completely random and there are certain patterns to be discover, AI frees human beings from amounts of data and helps find hidden models or factors from observed data (Boden, 2016).

However, AI practitioners still chase for general intelligence as the Holy Grail and hope that artificial intelligence can become general intelligence, namely artificial general intelligence (AGI) (Boden, 2016). AI is developing all the time and the twenty-first century is seeing the revival of interest in AGI with the recent increases in computer power. People hopes that with AGI, AI systems can benefit from even language, creativity, and emotion, which is still a major challenge. Take AI journalism as an example, one of the significant disadvantages of computer-generated news is its low versatility. The current technology cannot make it easier for the algorithms to learn about all fields and its role are necessarily limited to only specific domains with data pool and dry facts such as sports news and finance news. For other domains such as social news and entertainment news which need plot and descriptive words, artificial intelligence still cannot replace human beings since it lacks intelligence, creativity and human judgement compared to human journalists. Serendipity is a human gift, and human can come up with accidental inventions and discoveries with no reason. Without creativity, computer-generated content might sound technical and boring, while human journalists have the creativity to go beyond clichés and add humor to their stories.Besides, another one of the challenges is that AI programs don’t have a human’s sense of relevance, and the frame problem will still become a major obstacle for AGI as well.

Therefore, super computers aren’t enough while new problem-solving methods are often needed. Efficiency is important for this goal and problems must be tractable. In AI, Its Nature and Future, the author talks about several strategies for efficiency such as heuristic search, planning, mathematical simplification, and knowledge representation. Besides, artificial intelligence needs to learn and adapt to such changes in the environment and the system designers don’t need to foresee and provide the solutions for all possible situations. Artificial intelligence needs to predict changes in the environment and adapt their behavior automatically to better match the requirements of their task.


Alpaydin, E. (2016). Machine learning: the new AI. MIT Press.

Boden, M. A. (2016). AI: Its nature and future. Oxford University Press.

Warwick, K. (2013). Artificial intelligence: the basics. Routledge.