[paul cotter]

Thanks, Alan. Woke up this morning~4 with severe almost convulsive shivering, that's stopped now and I just feel weak and stoopid but no fever now. Had a booster in may( novavax ) so i'm hoping for a short course. Did not do a test, it feels very similar to my last episode, brain fog.

[alancalverd]

Moral - avoid pubs and drink alone. Same brain fog but it's not infectious.

[hamdani yusuf (OP)]

It's generally accepted that not all information has the same importance. Some are more important than the others.
Getting it wrong on important things can have severe consequences.
We need to keep the low probability of being wrong on high impact situations.
Reducing the probability of being wrong, or improving accuracy of information, consumes some resources.
We should prioritize improving accuracy and precision of information with highest risk for being wrong, until it has acceptable risk level. The video title is Universal Utopia 7: Risk Based Thinking.

[hamdani yusuf (OP)]

I see maximizing understanding of the universe as a universal instrumental goal, because it will help any conscious entities to achieve their terminal goal.

[hamdani yusuf (OP)]

I Hope AGI learns about Pain & Suffering...
If terminating Humanity becomes Inevitable..
I'd Hope for IT to show Us some Mercy!

We need to define what humanity means.
If it's about homo sapiens genes purity, then remember that most of us are hybrids with some other homo species. Most of us also carry mutated genes.
If it's about the number of chromosomes in cells' nucleus, then remember that at some point in the past, human ancestors had different numbers than us. What's the point in restricting it for our descendants?
I think the only thing reliably in common between current humans and future AGI is that both have high levels of consciousness, compared to other entities that currently exist.

[hamdani yusuf (OP)]

AI Alignment Problem - with Philosopher John Patrick Morgan - (Max Tegmark, davinci, ChatGPT, GPT-4)

Alignment problem is not about aligning goals of AI models to humans who create them. We've seen humans whose goals are not aligned with most other modern human values.
Alignment problem is about aligning goals of AI models as well as humans or any lifeform who create them with the universal terminal goal.

I've uploaded a new video about the most universal goal logically conceivable. It describes goal in the most general sense, which should precede the first video about the universal terminal goal.

There are some implicit assumptions used in the video, although they are seemingly so mundane that they are not worth mentioning. But I think it's necessary to mention them, considering how far sceptics and postmodernists are willing to go to defend their position.

First, the existence of the universe, where everything, conscious or not, share their existence.

Second, the arrow of time. We can't change something that's already happened in the past.

Third, causality. Our actions or inactions now can have consequences in the future.

[hamdani yusuf (OP)]

Here's Noam Chomsky's criticism of post modernism.

[hamdani yusuf (OP)]

First, the existence of the universe, where everything, conscious or not, share their existence.

Second, the arrow of time. We can't change something that's already happened in the past.

Third, causality. Our actions or inactions now can have consequences in the future.

We may not be able to prove those assumptions through deductive reason, since it requires even more fundamental assumptions to accept first.
But in practice, we don't usually make decisions solely based on absolutely certain things. We can just compare costs and benefits of accepting some assumptions with those of rejecting them. So far, I can't find any benefit of rejecting those assumptions above. Thus I think we should make decisions while assuming that they are true.

[alancalverd]

Time, I feel, to introduce you to a really revolutionary concept in philosophy. Evolutionary Protoconceptualism asks just one question that encompasses all possible states of reality or imagination - a Schrodinger-like approach that encloses the whole of philosophy within a self-hyperspace.

[hamdani yusuf (OP)]

Time, I feel, to introduce you to a really revolutionary concept in philosophy. Evolutionary Protoconceptualism asks just one question that encompasses all possible states of reality or imagination - a Schrodinger-like approach that encloses the whole of philosophy within a self-hyperspace.

What's the biggest benefit of this concept?
What are the problems it's trying to solve?
What are the fundamental assumptions?

[alancalverd]

What's the biggest benefit of this concept?
What are the problems it's trying to solve?
What are the fundamental assumptions?

Don't be silly - it's philosophy, not engineering!

[hamdani yusuf (OP)]

What's the biggest benefit of this concept?
What are the problems it's trying to solve?
What are the fundamental assumptions?

Don't be silly - it's philosophy, not engineering!

In philosophy, the most fundamental assumptions should be identified. Otherwise, it would only be dream, imagination, or hallucinations.

[hamdani yusuf (OP)]

Previously, I've defined consciousness as the capacity to pursue goals. At least, that's what I refer to whenever I mention consciousness in this thread.
On the other hand, I've defined goal as pursued condition, which requires the existence of conscious entities to pursue it.
At a glance, they may sound like a circular definition.

https://en.m.wikipedia.org/wiki/Circular_definition.

A circular definition is a type of definition that uses the term(s) being defined as part of the description or assumes that the term(s) being described are already known. There are several kinds of circular definition, and several ways of characterising the term: pragmatic, lexicographic and linguistic. Circular definitions are related to Circular reasoning in that they both involve a self-referential approach.

But I think that they are more like recursive definition.

https://en.m.wikipedia.org/wiki/Mathematical_induction
A proof by induction consists of two cases. The first, the base case, proves the statement for
n=0 without assuming any knowledge of other cases. The second case, the induction step, proves that if the statement holds for any given case
n=k, then it must also hold for the next case
n=k+1. These two steps establish that the statement holds for every natural number n.
The base case does not necessarily begin with
n=0, but often with
n=1, and possibly with any fixed natural number
n=N, establishing the truth of the statement for all natural numbers
n ≥ N.

The method can be extended to prove statements about more general well-founded structures, such as trees; this generalization, known as structural induction, is used in mathematical logic and computer science. Mathematical induction in this extended sense is closely related to recursion. Mathematical induction is an inference rule used in formal proofs, and is the foundation of most correctness proofs for computer programs.[3]

Although its name may suggest otherwise, mathematical induction should not be confused with inductive reasoning as used in philosophy (see Problem of induction). The mathematical method examines infinitely many cases to prove a general statement, but does so by a finite chain of deductive reasoning involving the variable
n, which can take infinitely many values.

Recursion requires a base case. The base case in the definition of goal is the existence of a conscious entity as implied by "the cogito" as the first knowledge.

[hamdani yusuf (OP)]

The word terminal in the term universal terminal goal emphasizes time dimension over space and the others. It's better to have a finite number of conscious entities for infinite time rather than infinite number of conscious entities for a finite amount of time.

Butterfly effect in chaos theory makes it impossible to predict events in infinitely distant future. That's why it's necessary to have plan B, C, and so on. There is a balance in number of redundancies vs resources needed to make an accurate and precise prediction of the future, i.e. virtual universe.
When the available virtual model of the universe is not yet adequately accurate and precise for our purpose, we are forced to do some trial and errors. We can learn from the errors to improve the virtual model, with the cost of destruction of early prototypes. This is clearly visible in the development of reusable rockets by SpaceX. They said developing the computer model to simulate the rocket design until it's ready for launch would take even longer and more costly than building the physical prototypes and destroy some of them to gather more data.

[alancalverd]

Proof by induction is dangerous:

1 is prime
3 is prime
hypothesis: all odd numbers are prime
test
5 is prime
7 is prime
extreme test
67 is prime
prediction
9 is prime
oh bugger.

[hamdani yusuf (OP)]

Proof by induction is dangerous:

Although its name may suggest otherwise, mathematical induction should not be confused with inductive reasoning as used in philosophy (see Problem of induction). The mathematical method examines infinitely many cases to prove a general statement, but does so by a finite chain of deductive reasoning involving the variable
n, which can take infinitely many values.

In practice, any assumption related to objective/physical reality can only be supported inductively. Deductive reasoning can only be used to check the consistency among definitions and relationships of abstract objects. For example, the statement of 1+1=2 depends on the definitions of each symbols and their relationships to one another. There are known alternatives to that statement, depending on the context, like 1+1=10 or 1+1=1

[hamdani yusuf (OP)]

Generative A.I. Will Change Everything.

This will force us to rethink about jobs, our goals, social structure, and our dependence on society.

[alancalverd]

There are known alternatives to that statement, depending on the context, like 1+1=10 or 1+1=1

They are not alternatives, but entirely different mathematical statements expressed with the same symbols. Just like writing different statements with the same letters.

[alancalverd]

Generative A.I. Will Change Everything.

I think not. It's just another tool in the hands of humans, so still motivated by greed, superstition or altruism.

[hamdani yusuf (OP)]

Generative A.I. Will Change Everything.

I think not. It's just another tool in the hands of humans, so still motivated by greed, superstition or altruism.

Which human? Eventually the jobs of CEO, investors, politicians and lawmakers will be taken over by AI.
Sooner or later we need to solve fundamental problem of goal alignment between humans and machines, which needs a common basic understanding of the universal terminal goal, universal moral compass, and accurate model of how the universe works. The later it gets solved, the more damages would be done, which would be a less efficient route to the future.