Why science based on induction is flawed

Isn't this deductive reasoning?

P1: If A was true everywhere we observed, A is true everywhere

P2: A was true everywhere we observed.

C: A is true everywhere

It's bad reasoning, but deductive reasoning is only as good as it's premises.

Deductive reasoning is only one half of the equation. You can have perfectly valid deductive reasoning which is at same time completely absurd. Let me demonstrate:

P1. If flowers exist, then I am God

P2. Flowers exist

C. I am God

Now, this is perfectly valid deductive reasoning. Conclusion naturally flows from premises. Yet premise 1 is obviously unproved, making this deductive argument not proved true.

However, moving from specific to general is not deduction. Its how induction is defined. You may call it deductive argument, but its still definition of induction.

Proper deduction moves from general to specific. When it moves from specific to general, thats induction or pattern generalization.

15 what?

Debby sometimes hallucinates. Happens to a lot of AI models.

If A does not conform to A then it is no longer A.

Yes, that is the most common critique of induction. Its that it doesnt follow law of identity (A = A).

While sometimes there is a strong induction, such as A which repeats everywhere and each time in same way, such induction only belongs to those hard science categories, and even there its not always consistent.

Induction is basically assuming that if A repeats here, it will repeat elsewhere too.

Sort of like, if truck moves 5 meters a second, it will travel 50 meters in 10 seconds. this is assumed true everywhere on Earth, and for a long time it was considered universally true until concept of wrapping spacetime, warp drive (from Star trek) and alcubierre drive came to mind, where something which normally travels 5 meters per second can at same time travel faster than that if spacetime is wrapped. Its sort of like making path shorter instead of increasing movement speed.

My point is the issue isn't unique to inductive reasoning

Of course, but it still remains an issue.

Also you will always have unproved premises in order to avoid circular reasoning or an infinite chain.

this is where I disagree. If starting premise is based on laws of logic, or on point that violations of laws of logic are a fallacy, starting premise is always true to anyone who accepts that logic exists and is a foundation of everything.

the only way to disagree with starting premise which is logical law,

such as:

A = A,

A = / = Not A (Law of non-contradiction, which foundation for contradiction is derived from),

A true under all conditions = A true under specific condition,

A true in all places = A true in specific place,

A true at all times = A true in specific time

is to abandon logic itself. I am comfortable if my debating opponent says he abandons logic.

As I said, the problem with induction is that it takes a different path from established logical law on deduction.

Established logical law of deduction is:

"A true in all conditions = A true in specific conditions"

or

"A true in each part of B = A true in specific part of B"

Socrates is human. (A is specific part of B)

All humans are mortal. (all parts of B are mortal)

Socrates is mortal. (specific part of B is mortal)

this is general to specific.

Induction is the other way around:

"A true under specific conditions = A true in all conditions"

Your post is specifically blaming the issue on inductive logic even though we both agree that the issue is with the argument and not the logic used.

Inductive logic is defined as an argument, the specific argument which I said is flawed. If issue is with the argument, its also issue with inductive logic defined by it.

"since we can convert inductive logic to deductive logic without changing the argument, we can do the same there."

Converting inductive logic to deductive logic does nothing to solve problem. Deductive logic which uses inductive logic (specific to general) will always face same issue.

I think you clearly agree with this, since your posts are based on examples and not collections of syllogisms. Heck, your initial post is almost entirely inductive logic.

No, my posts are deductive logic, not inductive. It may seem like "inductive logic" because I used examples, but examples dont make something inductive logic.

If I tried to prove the claim "all cats are black" as false, I would merely need one example of a cat which is not black. this is not "specific to general" inductive logic. this is "specific negating general" deductive logic.

Sort of like:

P1. All A are B.

P2. Some A are not B

One of these two must be false, by law of identity.

If I really needed to put my argument in pure deduction form without observations, it is that:

"A true in specific conditions = A true in all conditions" is an unproved premise which doesnt result from law of identity.

Its like saying "Some A are B = All A are B".

Conclusion doesnt follow, even if we dont use all the obvious examples where this specific logic is blatantly proved false. Law of identity simply cannot prove it, where in opposite case:

"All A are B = Some A are B"

Conclusion follows by law of identity.

to be clear, "=" in this case is forward deductive chain. It doesnt go backwards. A = B isnt same as B = A.

A lot of people confuse induction for observation, but the two arent same thing. Definitions, induction, observation, deduction are all separate forms of logic.

Observation would be:

"All cats I saw so far are black."

Induction would be:

"All cats I saw so far are black, so all cats are black"

Definition would be:

"I defined cats as black, so all cats must be black by definition"

Deduction would be:

"All cats I saw so far were black, so one cat I also saw before must be black."

Out of all of these, induction is weakest. Definition is at least true by definition. However, induction cannot even be proved by induction itself, or anything else.

I think you are conflating induction with just bad argumentation/fallacies

No, the basic rule of induction is:

"A is observed to be true everywhere, so it is true everywhere (even in non-observed areas."

or time based:

"A is observed to be true in all times, so it is true in all times (even in non-observed times)."

While large scale induction does increase probability with more observations being made, probability which you talk about is strong specific time/place based induction.

Its like me saying "Sun will rise tomorrow". I have seen Sun rise in this place each day for thousands of days. there is a chance Sun might not rise tomorrow, there is a chance that one day Sun wont rise (and that day will happen too, science says) but I expect it will rise tomorrow because constant repetition increases probability of something repeating. It is a probabilistic based reasoning.

However, you miss the point which I was making, which wasnt about specific time/place induction.

I was talking about all time/all place induction.

Making claims about all of the unobserved world is not same as making claims about specific area/time.

Maybe guy who only saw black cats does live in an area where all cats are black. However, if he were to make induction that all cats are black, that would be clearly false, even if he never saw a non-black cat.

Science faces these same issues when making claims about all of unobserved world based on observed world. Generally, the greater the unobserved world is, higher the mistake rate.

But the principle itself is flawed regardless.

Saying "A is true in observed world = A is true in non-observed world" is a leap of faith.

Forecasters sometimes get things right. Sometimes they dont. Some who use effective deductive reasoning, using principles which always repeat and which are expected to repeat in specific area, are one thing. Its sort of like betting that Sun will rise tomorrow, when you already know it did rise each previous day. Making absolute claims about whole of universe based on tiny portion of universe? Entirely different area.

Maybe large scale induction increases probability of being correct in a specific area and specific time, but even that faces many issues, as we see in psychology and sociology which are soft induction based science which often get things wrong. Any time you see an article saying "5 habits of smart people, according to psychology", its really just probability based. It doesnt mean someone who doesnt have those habits is not smart, or that someone who has them is.

Those aren't rules of induction, those are rules of bad induction.

And what are rules of good induction? Do tell.

Making reasonable assumptions based on intuition. Like if you travel across the country and look online and everyone you see is cheering for a specific candidate and he won the last 10 elections in a row, he's more likely to win the election than the guy everyone says they hate.

that is probability based induction. I agree some induction has higher probability than other, mainly the one which repeats more and where place is same and time is not distant. Same claim about that man across 40 years future prediction would be less likely, as well as that man trying his chances in another country.

It doesnt really solve problem of science in cases where 99% of existence is not observed, because observed repetition compared to non-observed world is lower.

Inductive reasoning doesn't yield absolute conclusions, and that is a strength of science not a flaw.

It is actually a flaw, because a lot of people get wrong conclusions due to induction.

"I was winning in gambling so far, so I will keep winning."

"I was safe when speeding in traffic so far, so I will keep being safe."

"I am smoking a lot and my lung health is great, so it will keep being great."

the problem with induction is in fact that its conclusion is not certain, so basing your thinking on that conclusion makes your thinking uncertain too. In simple terms, there is a decently high chance the conclusion is wrong. Even if that chance is merely 40%, it does make 40% of your conclusions wrong. And this chance of being wrong skyrockets the bigger the unobserved world is, and longer the unobserved time is. Its why I can make prediction that I will be alive tomorrow and have a reasonably high chance of being right, but same applied to next 40 years becomes less likely to be true, and we know this by observation of average population. However, when we dont know something based on observation even approximately, the chances of induction being wrong skyrocket. It becomes more like tossing a coin than doing probability predictions.

I buy the idea that AI is indeed doom for some people.

What I also realized is that a lot of people who oppose AI do so out of personal reasons.

AI didn't affect everyone positively.

Take for example, digital arts. Ever since AI appeared, these guys lost 30% of their income because why would someone pay someone to make digital art, when AI will do it for free.

Entire degrees in digital art went to waste, and people who today buy digital art are only doing so out of charity.

Today, if you want to be a digital artist, you cant be average. AI outperforms average. You need to be like way better than AI to merely survive.

Same applies to some other arts too. In past, people used to pay an artist to draw them. Today, you can just give photo to AI and have AI do it in few seconds, for free.

Artists are probably most negatively affected by AI out of all groups.

There are some other groups which were also negatively affected. Private instructors, such as instructors in English language, are also slowly being replaced. Advisors likewise.

Then there are also some people who were victim of AI, especially due to fact that AI can make porn out of anyone.

Then there is the predicted deepfake chaos. AI is being used to fabricate things, like AI being used to make fake photo of dead Benjamin Netanyahu. There are also social media accounts ran entirely by AI. Literally fake people and personalities on social media.

However, the main argument about AI taking jobs is that in past, machines used to replace unskilled labor, so unskilled workers got fired.

What's special about AI, is, that apparently, AI replaces both skilled and unskilled labor.

So there is I guess some fear that AI will replace everyone, because in future, there won't be a job AI can do.

However, I think people who make these arguments about jobs don't realize this isn't a bad thing. Having an AI worker to do job for me, actually sounds pretty good.

Creation of AI was really like opening a Pandora's box. Once opened, it cannot be closed. Everyone knows how to make AI. There is no way to effectively regulate or stop it.

So I am afraid that doomsayers better get used to AI, because it's not going to go away. Actually, from this point on, the only question will be who has a better supercomputer to host a better AI.

The idea that boycotting AI is gonna stop it, is silly at best.

AI is pretty much here to stay, and this is true even if all countries banned it.

Induction works in scenarios where uncertainty is built in

It works sometimes, sure. My complaining is more about time when it doesnt work, which happens a lot. My issue is that probability as a way to reduce uncertainty only works if we assume unobserved will follow same probability as observed. It's a huge leap of faith in all cases where amount of observed data is insufficient or just tiny.

I mean, how exactly do you use deductive reasoning to determine what the next card will be in a shuffled deck?

Usually by knowing which cards aren't in the deck and which groups of cards are more present, so the ones which are have higher probability of being drawn.

Its sort of like if 50% of cards were white and 50% black. If you drew 6 cards and all were white, the remaining cards contain more black cards than white cards, so higher chance of drawing black card.

Those who play poker usually divide cards into good cards and bad cards, and if you have a really good hand, there is actually a higher chance opponent doesn't have even average hand. But same works in opposite. If you have below average hand, chances are higher that opponent has above average hand.

I haven't played poker in a long time, so I can't really remember all the tricks. Math changes based on number of players, because more players means higher chance someone has a better hand than you do.