Hacker News new | past | comments | ask | show | jobs | submit login

Yes, I often wonder too where all these new inventions are gone after the initial enthusiasm and media attention fades. Probably every weeks you can find articles about 'breakthrough' in material science, battery, memory technologies etc. Yet when I try to Google them after few years, >90% has no follow ups nor these technologies reach market. I guess this is like startups -- good idea is not enough, execution and financing is what matters.



To an extent, also a lot of these processes are incremental rather than revolutionary (despite the press releases) and do eventually make it through to production after an upgrade cycle or two.

Hell a carbon fiber (partially) plane flew around the world non-stop in 1986 but it wasn't until 2012 that carbon fiber road bikes got good enough/cheap enough for me to buy one.

I think people underestimate incremental improvements, 4% (arbitrarily) a year, year on year adds up eventually.

Look at the improvements to Li-Ion battery technology as a good example.

Recently I've been doing research on a programming problem that has come up at the new job, I pulled the related research and the canonical first representation of the problem on a computer was formulated in 1966 and the issues it addressed in the paper are identical to the way the processes are run at new employer, almost word for word.

51 years that research has existed and they are still doing things the way they did then.


> Recently I've been doing research on a programming problem that has come up at the new job, I pulled the related research and the canonical first representation of the problem on a computer was formulated in 1966 and the issues it addressed in the paper are identical to the way the processes are run at new employer, almost word for word.

Computing is weirdly ahistorical. Hardly anyone ever looks at the research; everyone prefers to invent it themselves.


Absolutely, doing a deep dive into the material has been fascinating, you can see the approaches evolve as hardware got faster.

My math level isn't quite there for some of it but the problem had been studied hugely and there are some good 'field guide' level references out there.

Its made me consider going back into education to do maths though. I don't like that I don't grok everything and with practical applications I'm actually excited by the maths.


> Its made me consider going back into education to do maths though... I'm actually excited by the maths.

Warning: One of two things will probably happen if you go back to (grad) school for math. Either you'll lose enthusiasm in the first year or two because you really love building stuff and miss it, or else you'll find you really do love the math and spend the rest of your life at a blackboard :)


What field of CS is this, if you don't mind me asking?


Flow shop scheduling, its a bunch of different approaches to optimising a seemingly simple problem (at first glance) that turns out to be np-hard at second glance.

The more I read the greater the complexity, I've never been much on the theory side and frankly as an enterprise programmer I've never really had to be, my distant A-level math has always been enough.

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

Its a practical application of some really beautiful approaches to a problem (everything from simple queue stuff through to genetic algorthithms and machine learning) I didn't know existed and at the same time a decent solution will have a really big impact on the business I work for.

It might take me quite a while to grasp even a small chunk though I'm starting from a pretty low level.


I'm not the parent poster, but for me it's confidential computing. I'm in the same boat, though I do have a B.Sci in Mathematics which helps!


articles always overhype these types of advances. Specifically in materials science, they ignore:

1. ease of synthesis. Many of the cool materials with nice properties you read about in these articles cannot yet be produced easily at scale, at requisite purity, and cheaply. Tying in to...

2. cost. Even when new materials are strictly better than widely used ones, they can still go nowhere. Obvious but everyone seems to forget this.


Yes. I feel like cancer has been cured every few weeks since Reddit first went live.


Cancer is easy to kill. The hard part is not killing the human around the cancer as well.


Add to this that there are many, many different types of cancer. There won't ever be a cure for cancer. We may find a cure for a cancer, but there are many types of cancer.

A good friend is an oncologist. He kinda hates it when people talk about a cure for cancer. I guess it just doesn't work that way.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: