11 Oct
2024
11 Oct
'24
10:25 p.m.
Alexandre Rademaker: We don’t necessarily need to throw away the meanings. A safe
translation should account for a 1-N mapping.. from surface to logical representations.
Context or even some statistical preference can select the most preferable reading.
Yes. That is why we need a top-level symbolic processor that can determine what to do for
any particular issue that may arise.
Alex Shkotin: With robots it's better not to use vague terms or sentences. It's
dangerous. Good robots will tell: I don't understand, bad ones can make a mess of
things.
As I said to Alexandre, the top-level processor should use symbolic methods for
determining what to do.
Alex: My way is to represent knowledge formally. The precision of knowledge itself
remains the same initially and may be better after we apply knowledge processing
algorithms to this formalized knowledge.
Think of the top-level symbolic processor as a gate-keeper. It is in the best position to
determine what to do. In many cases, the best thing is to ask a question or even a series
of questions before making a decision.
The top-level processor may use LLMs in the simplest and most secure way: Translate a
query in any natural language to and from whatever internal form the system uses. After
the top-level processor has determined what to do, it can pass the translated result to
whatever subroutines can handle it. Those subroutines may or may not use LLMs or many,
many other tools of various kinds.
Basic point: One size does not fit all. The top-level processor determines which of many
internal processors should or should not be invoked. Anything that seems dangerous can be
sent to a security system, which may or may not reject the input or even send it to proper
authorities to handle it.
John