What would be a general definition for intelligence and cognition?

[ALine]

What would be a formal definition of intelligence/cognition. Which that can be expressed formally and mathematically?

One which could be ran according to unique algorithms/computation/mathematical methods.

[iNow]

What type of intelligence?

 

[ALine]

4 hours ago, iNow said:

What type of intelligence?

 

I mean like a formal definition of intelligence/cognition. One where it can be expressed by a computer.

[iNow]

I’m sorry. I have no idea what you’re talking about, and I believe the core explanatory reason is because neither do you. 

[dimreepr]

7 hours ago, ALine said:

I mean like a formal definition of intelligence/cognition. One where it can be expressed by a computer.

The closest I can come to a formal definition is a 'joke', one form of intelligence makes a joke for another form of intelligence to laugh, if they don't laugh they're not cognisant of the intention, a computer won't ever laugh, however intelligent we make them.

[Ghideon]

Your question is complex and contains unresolved issues and active research.

On 5/15/2024 at 9:16 PM, ALine said:

What would be a formal definition of intelligence/cognition.

That is ar far as I know an unresolved question; the scientific community does not have a single, universally accepted formal definition of intelligence.

On 5/15/2024 at 9:16 PM, ALine said:

Which that can be expressed formally and mathematically?

Then we have to assume that intelligence* and cognition can be expressed mathematically in a way that is useful. There are attempts in for instance Computational Models of Intelligence and I am not aware of any consensus. You may want to look for work that uses Kolmogorov complexity and Markov blankets (I do not have any sources; tried and failed to locate an article I read some time ago)

On 5/15/2024 at 9:16 PM, ALine said:

One which could be ran according to unique algorithms/computation/mathematical methods.

That assumes that Intelligence can be expressed in a model of computation. As far as I know the debate about whether all aspects of what we consider "intelligence" can be fully captured and replicated by computational systems is not settled. 

Examples to illustrate the complexity: You might be able to mathematically define some type of intelligence that performs well on a typical IQ test**. That does not mean the same mathematical definition is applicable to the intelligence required to :
-construct new IQ tests
-evaluate results of IQ tests
-Ride a bicycle to the facilities providing the IQ test.
-decide if it is appropriate to use an IQ test in a certain context; moral or ethical. 
(These examples are inspired by iNow's answer)

 

*) Or intelligences, as @iNow correctly points out.
**) such as those used by Mensa or other organisations 

[ALine]

3 minutes ago, Ghideon said:

That assumes that Intelligence can be expressed in a model of computation.

I am assuming that it can be.

4 minutes ago, Ghideon said:

Then we have to assume that intelligence* and cognition can be expressed mathematically in a way that is useful. There are attempts in for instance Computational Models of Intelligence and I am not aware of any consensus. You may want to look for work that uses Kolmogorov complexity and Markov blankets (I do not have any sources; tried and failed to locate an article I read some time ago)

Thank you, looking into it now

[dimreepr]

14 hours ago, ALine said:

I am assuming that it can be.

Assuming anything is the wrong place to start, if you want the truth. 

What would be the benefit of a computational model?

For instance, would the joke be funnier?

The computer say's "the joke is n% funnier, when humans listen to it in a group".

What does that tell you about the joke?