Very short written about true and usefulness in sciences and engineering

Introduction

From a long time I have always thought that people from STEAM careers ignore a lot about the philosophical roots of their fields. Even very fundamental concepts about true is given by understood. I do not think this is necessarily wrong, but if not taken carefully, but in some cases I think it becomes harmful. So I want to write a little bit about what true is in science field and philosophical field. And I will try to explain why I do think it is an important matter to know.

Why humans seek answers

Human beings have always funded interest in answer question, because it has been our way to survive to nature. From knowing what food is fine to eat, when can you plant seeds to get a good harvest, why can you get a material to construct something and many other activities are subject to get an answer. Ancient people found solutions to a large amount of matter without a carefully designed experiments not even with a systematic mechanism of thought, however they found solutions to a lot of their problems. But to find a solution to a problem is not necessary to find a true answer. For example, Goldbatch's conjecture says that any even number greater than 2 can be expressed by sum of two prime numbers. At this time, that problem has no solution. No one knows if that conjecture is true or false, but if have been proved true up to 4 quintillions!. Ancient people could have been discovering or doing up to 4 quintillions of cases or experiments to a problem and then to think their solution or hypothesis was true, but it was not. That is why philosopher has tried from a long time to establish a mechanism of thought to distinguish what is the true and what is not. In summary, non solution have been found to this problem, but there are excellent approaches.

Solutions versus truth

There has been large debate among philosophers about what is the true and if it can be understood: skepticism, pragmatism, postmodernism, etc. are some philosophical stance that tries to reply those interrogate. But to start with a proper study of the phenomena we will assume that true exists and that true can be understood for us humans. In contrary case, then no discussion matters because there will not be anything to prove. Now, the other theme is where and how true needs to found. Both questions are different and important. Where asks for the plane of reality where true takes place: nature, mind or something else; while How asks for the mechanism that us humans need to find the truth.

For the fist question one possible answer is that true comes for the plane of ideas, and not only from nature. There are concepts and natural phenomena that need hypothesis or theory frames to be understood. Those concepts reside in the plane of ideas, and plane of ideas is different from the plane of mind, because I have my mind, with my belief, my fears and misconceptions; and you have your mind, with all your belief, fears and misconceptions; but ideas do not have them. There are a set of significant that describe something that we humans can think about. But even if we die, even if humans and other intelligent creature cease to exist, ideas still persists. A triangle will always be a polygon with tree sides, even if universe no longer exists. So with that arrives the issue of how can we access that true, how can we access to the plane of ideas.

Mathematics and proof

One of the few science fields that has been in the pedestal of being objectively true is mathematics, because in this field the object of study and its significant can be very well-defined. So with a proper mechanism usually called mathematical logic they can be proved (and still mathematics has some few possible cases of contradictions, like the one caused Cantor's set theory). The reason for this is that mathematics studies starts from axioms and proves conclusions with a deductive process. In that way deductive reasoning is how we access the plane of ideas (at least to those that are not "autoevident" unlike axioms). But this is not the case of sciences fields and philosophy. For the case of sciences the reasoning is done by induction: first reality is observed, a model is build and then test are conducted to make predictions. The process is repeated indefinitely. In this case induction is how we try to get a true. But induction is not completely different from try to prove Goldbatch's conjecture by proving all numbers; it is impossible. It is very hard if not impossible to define what is gravity, it is only something that we humans can sense, we know that it exist but how to define and explain it not easy even if measurable with high precision. That is why by ages the most accepted description of gravity came from Newton's mechanics, but then Einstein showed the limits of Newtonian mechanics and produced a theory with a broader range of validity. For the case of philosophy reasoning may be inductive and deductive, but philosophy also deals with matters that cannot be subject of physic experiments, since it also deals with questions that are not true or false like "Why science is no longer called natural philosophy?" or "what is the importance of culture?", and it also deals whit definitions and concepts like "what is the difference between science, mathematics, and philosophy?" and "what is a democracy?".

Science and falsifiability

Ideas has a set of significant that give "form" to something we can think about. So, if it is possible to us prove that a set of propositions leads necessarily to a conclusion then we know true. And the value of true can only be known by the significance of each proposition. And significant cannot have contradictions. For example, if I say that a prime number is only entirely divisible by 1 or itself then to prove that a number is prime the only way to prove it is that no other number except 1 or itself can divide it. That way the definitions contains no contradictions and the true value can be known if I know what is the meaning of "entirely divisible by 1 and itself". And this leads us to a conclusion. True can only be proved if there is no room to doubt of contradiction. In contrary case, I may only have a "good" hypothesis that has been proved true for many cases. In addition, a "good" hypothesis or a hypothesis that can be subject of study of true value is only one that have a criterion which allow to be proved wrong with a false case. This is called falsifiability. Worth to mention that falsifiability do not prove true, instead is a criterion for theories, since falsification can eliminate bad theories. Theories that survive along attempt to prove them wrong gain "grades of true". This is what true is in science and engineering: theories or models with high grades of true, that allow predictions of high precision.

Engineering and usefulness

So even when science has been proved valuable for society because of knowledge it has given us, that it is used in technological advances, microelectronics, software, medicine, agriculture, etc. The knowledge cannot be taken as written in stones, as something that cannot be denied, that mindset is contrary to someone who cares to seek the truth. Even more these are just consequences of predictive capacity on scientific theories, but they are not true, they are just useful.

Engineer students often have to ask questions: "Does this work? Is it Safe? Is it efficient? Is it economic?" and they sometimes answer those questions in functions of pure usefulness and mathematical theories, but they do not care about the true behind those theories and that "usefulness"... Useful to what?. In some philosophical perspectives, this usefulness is view as the true itself; as in pragmatism. Not only that, but sometimes, they are given a moral value like scientism. This philosophical views have also been inserted in conception of technological advances and their results. In consequence the answer to the questions that Engineer students give comes involved into philosophical views with a set of repercussions in real world if not taken carefully. One consequence of equating usefulness with truth is that technological progress itself can begin to acquire moral authority. For example, Transhumanism is an intellectual and philosophical movement advocating for the use of advanced technologies—like genetic engineering, AI, nanotechnology, and cybernetics—to radically enhance human physical, cognitive, and psychological capacities. Its ultimate goal is to overcome biological limitations like aging, disease, and mortality. And even when there are voices against transhumanism warning about its possible consequences, there are who just attack those voices saying that they are against science, if they are against science, then they are dumb, they cannot know what is true: true is that technology is always useful to humans, true is that science is not the seek of true, but it is an instrument to achieve the perfection of humans.

AI as a modern example

We Engineers are like those ancient people: We find solutions to a problems but not trues. And that is specially relevant in fields like machine learning and AI. AI only cares about relation between training data and expected results, based in the probability that the relation found fits well the expected result to the input. It does not care about if that relation is true or if it is not. It only cares about correlation, not causality. Furthermore, it cannot create knowledge, nor even generate information. Information is surprise. Surprise only arrives when something unexpected happens. AI only generates data, data that may be useful to someone. Data that is needed to take decisions. When that decision, becomes useful there is sense of true on it. When this happens frequently correlation and causality becomes the same in people minds.

Why this matter to us Engineers? Because engineers sometimes do not care to prove results. Every engineering model is known to be wrong, but many are useful. Sometimes they have taken by true, for example, the results of an IA, because many times previously it was useful. If that mindset is applied to a technology that may have real impact in humans then it becomes dangerous. For example, IA that classify CV to get a job could be biased to "think" that someone is not suitable because that person is a man, and if the IA was trained with CV from professionals majority women, creating a scenery of discrimination with serious impact of good living of people.

Why intellectual humility matters

To accept that you do not know everything. To accept that no one else know everything. To accept that there are trues that remain unknowns, because no one knows everything helps you to stay alert about consequences of actions. Scientific knowledge, technological advances, philosophical discussions are actions that produce several impacts in our history, in the direction society takes. So that, the seeks of true remains more important than ever, because as more dangerous "useful" technologies becomes, more important is to try to know what is true or the "grades of true" about the matter that are said around them.