Ravi,

Probability is another method for dealing with many kinds of continuous issues.  Fortunately, the mathematical methods of probability and statistics are very well developed.

This is another kind of symbolic reasoning that LLMs, by themselves, cannot handle.  A system that uses symbolic methods can invoke reasoning methods of many kinds:  formal logic, probability, statistics, and various computational tools.

Arithmetic, for example, is ideal for a computer, but LLMs are horrible for anything except trivial computations.  There is 60+ years of symbolic reasoning methods in AI and computer science.  LLMs can't replace them.

General principle:  Symbolic methods must be in control of the overall system.  They can determine which, when, and how other methods, including LLMs, can be used.  They can also prevent the dangers caused by runaway AI methods.

John
 


From: "Ravi Sharma" <drravisharma@gmail.com>

John
What happens to statistical entities which most are? If we can not define them by FOL what do we do?
I realize we can apply logic to artifacts (real) that are statistical in nature, for precise filtering etc. as example.

This brings me back to what tools are going to be available in future cyber or AI scenarios that would have some ability to understand context, provenance, real or virtual tagging etc so that we can distinguish real vs "processed' reality?
Thanks.
Ravi