Only because I don't yet see it mentioned, the one emerging field to rule them all: program synthesis. :P
Ask HN: What are the most interesting emerging fields in computer science?
61–70 of 183 posts
Re: Ask HN: What are the most interesting emerging fields in computer science?
#62Deep Video Portraits - https://web.stanford.edu/~zollhoef/papers/SG2018_DeepVideo/p...
DeepFake Creation Tools - https://voat.co/v/DeepFake/2405562
Adobe Voco = https://en.wikipedia.org/wiki/Adobe_Voco
Re: Ask HN: What are the most interesting emerging fields in computer science?
#63Re: Ask HN: What are the most interesting emerging fields in computer science?
#64Any field that isn't being outsourced (yet).
Re: Ask HN: What are the most interesting emerging fields in computer science?
#65Computer Vision.
For computer vision without context (2D images standing alone), we have some nice solutions already, but I think that as long as we keep using the same methods, it will be insufficient for many purposes. Because the truth is that projection of images to a square, 2D grid, and given the complexity of lighting, put us at a situation where we have insufficient information.
And we are already seeing this being heavily developed in autonomous driving systems and others, but I feel like the biggest computer vision applications will require much more information than a 2d image can offer. Instead, recognising objects when you have 3d information seems much more reasonable to me.
Re: Ask HN: What are the most interesting emerging fields in computer science?
#66I think most interesting computer science fields are actually application of CS in other domains. Science changed a lot in the last decades, moving from a genius in a room looking at the data and coming up with grand theory to have vast amounts of data that no single human can make sense of. The work of the computer scientist is to quickly understand problems from various fields then solve it using tailor-made algori…
Agreed. CS is the study of translating human needs into formal systems. In a way, it's "applied philosophy".
I think taking CS and "bouncing it off" of other disciplines is where the real magic happens.
Re: Ask HN: What are the most interesting emerging fields in computer science?
#67Earlier quoted context omitted.
The biggest barrier to adoption of formal verification that I have seen as someone just starting in the field (working through Software Foundations and have a number of projects planned with SPARK, Frama-C, and LiquidHaskell) is the lack of groundwork. Verifying just your own code is complex enough as it is but working with libraries without any clear specification of their interfaces and behaviours makes this so muc…
To me, personally, the biggest barrier was lack of a proper introduction with a lot of examples. I try to break this barrier a bit with my upcoming book: Gentle Introduction to Dependent Types with Idris. I am very interested in this area but it is impossible for newcomers to get a grasp of it without too much digging. Logical Foundations was OK but I was still missing the theoretical explanation ("why does this tact…
Ooh I'll check this book out once I get through the pile of stuff I have right now.
And as you noted there definitely is a lot of "magic" when it comes to the inner workings of theorem proving tactics. I'm slowly figuring all of that out but like you said it definitely takes time and digging at the moment.
Re: Ask HN: What are the most interesting emerging fields in computer science?
#68Earlier quoted context omitted.
Rust looks like a step towards practical formal verification just because of its design philosophy. I think what we're doing is making engineering languages more and more verifiable as research languages become more and more expressive. Eventually they'll meet in the middle and we'll have formal verification in "real life."
No progress in language theory can fix the fundamental problem of software verification: you need a formal specification to have anything to verify. Who wants to write not only a detailed spec for their code, but a spec that has well defined semantics in some kind of logic? Nobody, that's who. There are very few properties that you care about that are both sufficiently easy to encode in a formal specification and not…
In the logical extreme that can't be the case, because a fully debugged program plus a machine to run it on actually satisfy the definition of a formal specification. Deciding what you want your program to do is the eternal burden of programming, but maybe there's a way to make formal specification at least as easy as regular programming.
Re: Ask HN: What are the most interesting emerging fields in computer science?
#69Earlier quoted context omitted.
Just another hype. Sensors and internet are here for decades. Internet is ok, but “things” are way too expensive and not reliable yet.
IMO the best IoT is the DIY IoT, the kind that isn't really IoT but rather "I put Wifi on a raspi and connected it to a PCB". Thankfully the online resources around electronics are plentiful and PCBs can be had for under 10$ incl. S&H. That way I can make all my lighting IoT without having to deal with the garbage of the IoT industry.
Yes, IoT is great when you can DIY. But when you need to send a technician 2-4 times a year for each node... it’s a problem, not a business.
Re: Ask HN: What are the most interesting emerging fields in computer science?
#70Secure Multi-party Computation.
The basic idea is developing methods for two (or more) parties with sensitive data to be able to compute some function of their data without having to reveal the data to one another.
The classic example is developing an algorithm that allows two people to figure out who is paid more without either revealing what their salary is.
Such algorithms get significantly more complicated if the threat model starts changing from "we're all acting in good faith, but we just don't want to share this private info" to "I'm not sure some of the people involved in this are acting in good faith."
Based on my (admittedly limited) look into this field, it seems like there has been some theoretical progress made here, but there's nothing like a generalized framework or library for general development with it. Instead, practical applications seem to be one-offs. For example, a contractor a while back developed a system that lets parties (nation-states or private space firms) figure out if their satellites are going to run into each other without revealing anything about the location or orbit of their satellites. That way they don't share sensitive data, but they can move their satellites if they're on a collision course with somebody else.
Personally, I got interested in this when working for the government. I was working on an extremely cool data integration project (State Longitudinal Data System grant form US Department of Education) that basically went nowhere because we couldn't get over the legal hurdles to data sharing... If we didn't have to share data, but could still compute interesting statistics about the data, that would have been really cool.