The SP Theory of Intelligence, and Its Realisation in the SP Computer Model, as a Foundation for the Development of Artificial General Intelligence

J. Gerard Wolff

DOI: https://doi.org/10.3390/analytics2010010

2023-02-17

Analytics

Abstract:The theme of this paper is that the SP Theory of Intelligence (SPTI), and its realisation in the SP Computer Model, is a promising foundation for the development of artificial intelligence at the level of people or higher, also known as ‘artificial general intelligence’ (AGI). The SPTI, and alternatives to the SPTI chosen to be representative of potential foundations for the development of AGI, are considered and compared. A key principle in the SPTI and its development is the importance of information compression (IC) in human learning, perception, and cognition. More specifically, IC in the SPTI is achieved via the powerful concept of SP-multiple-alignment, the key to the versatility of the SPTI in diverse aspects of intelligence, and thus a favourable combination of Simplicity with descriptive and explanatory Power. Since there are many uncertainties between where we are now and, far into the future, anything that might qualify as an AGI, a multi-pronged attack on the problem is needed. The SPTI qualifies as the basis for one of those prongs. Although it will take time to achieve AGI, there is potential along the road for many useful benefits and applications of the research.

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is: **Whether the SP intelligence theory (SPTI) and its implementation in the SP computer model can serve as a solid foundation for the development of artificial general intelligence (AGI)**. The author explores this issue through the comparison and evaluation of SPTI with six other alternatives (ALTs).

### Specific Objectives of the Paper

1. **Evaluate the potential of SPTI as a basis for AGI development**:
- The author believes that information compression (IC) is an important principle in human learning, perception, and cognition, and SPTI realizes this principle through SP - multiple - alignment.
- SPTI not only has descriptive and explanatory power, but also flexibility and simplicity, which makes it a promising basis for developing AGI.

2. **Compare with other potential AGI bases**:
- The paper details six alternatives (ALTs), including "Society of Mind" (SOM), DeepMind's Gato, OpenAI's DALL·E 2, Soar, ACT - R, and NARS.
- By evaluating these alternatives and SPTI in multiple dimensions (such as simplicity, the ability to model human intelligence, other advantages or disadvantages), the author shows the advantages of SPTI.

3. **Propose a multi - pronged research strategy**:
- Since the road to AGI is full of uncertainties, the author advocates a multi - pronged strategy, that is, exploring multiple possible research directions simultaneously.
- SPTI is regarded as one of the important research branches, although it is not the only solution.

### Conclusion

The main conclusion drawn by the paper is that **SPTI is indeed a relatively promising basis for AGI development**, especially performing well in information compression and SP - multiple - alignment. However, the author also emphasizes that achieving AGI will take time, and many useful applications and intermediate results may be generated during this process.

### Example of Formula

The information compression (IC) mentioned in the paper can be represented by the following formula:

\[
IC=\frac{\text{Original data volume}}{\text{Compressed data volume}}
\]

where \(IC\) represents the information compression ratio, reflecting the degree to which the original data is compressed into a smaller representation form by a certain algorithm or model.

### Summary

This paper explores the possibility of developing AGI through the comparison of SPTI and various alternatives, and proposes an intelligent theoretical framework based on the concepts of information compression and multi - alignment.