In the last half century, Artificial Intelligence (AI) has become a hot topic among philosophers and scientists. As we take AI for granted, some have argued that we are too far away from creating a sentient being. The answer depends on whether you view AI as a way to improve human life or an end in itself. To be able to answer this question, you need to understand some of the current challenges in AI research.
In the 1950s, the Rand Corporation, which had been a prominent place for AI research, engaged the philosopher Dreyfus to produce a critical report on AI. Though leaders of the organization argued against publishing Dreyfus’ report, it became the most-requested report in the history of the Rand Corporation. The report was later expanded into the book What Computers Can’t Do. Dreyfus argued that humans have a large proportion of tacit knowledge that cannot be incorporated into a computer program.
Despite all the hype around the creation of AGI, scientists are still far from reaching this goal. Some scientists believe that AGI will lead to a new phase of AI development: Artificial Super Intelligence. This would be an AI that surpasses human capabilities, from decision making and reasoning to making art and building emotional relationships. However, there is no concrete evidence to support this claim. Only time will tell. It’s still far away from becoming a reality, but the benefits are worth imagining.
The research behind AI continues to change rapidly, and advances in machine learning are nowhere near being close to AGI. But there is still progress in the field, with the most recent breakthroughs occurring in neural networks. Artificial neural networks mimic the brain through code. However, this form of AI is still a long way from the real thing. AGI has been feared by some people, but advances in artificial intelligence have paved the way for the development of this technology.
AI can predict the demand for products and services based on the data collected by employees. It can analyze employee data to predict when an employee is about to quit, and it can even interpret customer service queries. The benefits of AI are many: it can interpret video feeds from drones, help radiologists spot cancers, flag inappropriate content on the Internet, detect wear and tear in elevators, and generate 3D models of the world.
The differences between human and AI systems include basic structure, speed, connectivity, and scalability. Humans cannot directly communicate with an AI system, but they can indirectly through language and gestures. This method requires a limited bandwidth and is slow compared to the speed of AI systems. AI systems are also connected directly to each other and collaborate based on integrated algorithms. There is a minimal risk of miscommunication between the two.
In general, human beings are better than AI systems at social and cognitive tasks. In particular, humans are better at social-psychosocial interaction. However, AI systems are not yet capable of interpreting human language and symbolism. They also require a complex frame of reference that is far beyond what an AI system can do. These are all reasons why AI systems may never become as advanced as human beings. They may be decades away from being able to learn how to communicate with humans.
Several examples of AI progress include Google’s self-driving car, which has passed a state driving test. Another breakthrough was the defeat of world Go champion Lee Sedol by Google DeepMind’s AlphaGo. Go had been considered a major hurdle for AI until now. Moreover, Hanson Robotics’ Sophia robot is capable of facial recognition, verbal communication, and expression. A BERT system is also used to reduce barriers in understanding by machine learning applications.
As AI technology develops, it is important to understand its limitations and strengths. We should understand the underlying qualities and characteristics of AI systems and learn how they “think”. As the AI evolves in the future, its capabilities and autonomy will become more relevant to human decision-making. But until we learn how AI works and how to trust it, we can only use it for research. There are many societal challenges ahead, but we should not be discouraged.