Hacker News Reader

Comment by smackeyacky | Hacker News Reader

But science is objectively better than any alternative we have found so far.

Science works very well for things like not getting sick, understanding physical phenomena or getting to the moon. But for some goals like being happy, finding meaning in life or cooking a tasty meal science is only of limited use.

Also I don't think the author necessarily proposes to replace science, but rather being aware of the limits of science. There is no one scientific or rational way of coming up with scientific hypotheses or determining how to apply scientific discoveries best to the real world. There is always a gap between the ideal world of science and reality.

You appear to believe science can't be used to analyze subjective experiences such as happiness or a tasty meal? Consider this - science explains why cilantro tastes like soap to some people. A genetic test could determine if you're going to experience cilantro tasting like soap and therefore not like it. We're just starting to scratch the surface. We may very well be able to do a genetic test and accurately predict/prescribe your favorite meal. I imagine we'll make a lot of progress in this area over the next 100 years.

Science can be used to analyze subjective experiences, but analyzing something and being the one and only all-knowing oracle is a big difference. Science is always constrained to an ideal world, but reality is messy. There is no one correct way to pull out a piece of egg shell that got into the pan by mistake while cooking.

Also I believe predicting things like favorite meals is possible to a high degree of certainty, but there is always a gap between the nice theoretical world and the real world. Maybe it's your favorite meal because your grandmother always cooked it for you. How do you define a meal? How much egg can you add or remove before it becomes a different meal?

If my arguments seem too hand wavy, consider the incompleteness theorem. There are truths in mathematics that cannot be shown to be true. Why should messy real life give you stronger guarantees than the clean world of mathematics? Why should there always be one correct way of doing things that we can deduce?

Or let's say I bet that a certain Turing machine doesn't halt. What does science say about this? Is this rational? Should I have bet that it does halt?

Science is one of the best tools available, but it will not solve all of humanity's problems.

It's not unreasonable to assume that some of humanity's problems are NP-hard and so while science won't be able to solve those problems, neither will anything else. However, that's if we're looking for exact solutions, for we can get far with statistical modeling - we might even be able to get to good enough. I have a sneaky suspicion you would find such good enough solutions highly unsatisfying because I suspect you're after Truth, which may not be knowable, even in principle.

Reply on news.ycombinator.com