Artificial General Intelligence (AGI) is our next step towards near-human intelligence by advanced technologies. It’s not a specific set of intelligence such as chatbots, chess plays, image generations, or LLM, but in general sense will mimic human’s comprehensive cognitive abilities. Understanding, sensing, reasoning, adapting, solving, reacting, etc., will all be within the AGI’s capability list. Some used term “hypothetical” to discribe current imagination of AGI’s goals, but with the progress of technologies and sciences, AGI may become closer and closer to reality, faster than common population could realize.
From DeepSeek R1, it gave a summary of current stepping stones towards AGI: (Such research can be easily done by today’s AI)
- Large Language Models (LLMs)
- Multimodal AI – integrating multiple types of data and sensory data in models
- Robotics and Embodiment – interaction with the world to accelerate the learning of the robots
- Reinforcement Learning (RL)
- Neuromorphic Computing & Brian-Inspired AI – hardware architecture mimicking human brian neural structures
For the future advancement of AGI, huge challenges, and fear, coexist. To achieve the comprehensive abilities of the humans, both technology and science need to go hand-in-hand. A multi-disciplinary approach is a necessity, and not just data processing. From this sense, AGI could be a relatively distant goal with immense scientific and technical challenges -a slight relief sign. Ethical and social challenges surely need some serious debating and thinking. However, once collective humans set their minds and resources to a goal, it will always bring some amazement. It may bring us to the next age although may be fraught with worries and problems.
See our earlier article for more thoughts: With AI, Are We Really out of Hope as Humans?