« first day (930 days earlier)   

mdx
1:20 AM
I found the answer to a cs.se question from an equivalent question on cstheory.se (cs.stackexchange.com/questions/30216/exponential-analogue-of-nc). The cs.se post cannot be marked as a duplicate because cs.se != cstheory.se. I put the link as a comment. What's the appropriate way to resolve that question (on cs.se)? Having an identical answer to the one on cstheory.se feels inappropriate, but leaving it unresolved doesn't feel right either.
 
vzn
1:44 AM
was your intended answer going to be about the same as cstheory.se? your comment "resolves" it (nice find!). there is some overlap/ duplication across se's and comments are the main way to point this out. migrations are another mechanism but they are not used for duplicate questions. note that questions cant be migrated after [x] days. something like x=60.
 
 
1 hour later…
mdx
2:51 AM
It's hard to say where it was going since I wasn't completely done with it (I revise and edit my answers before posting a lot). It would have trailed towards Ryan's answer, done something with a strange notion of uniformity, or just end up being wrong, haha. Maybe I (or someone else) might try to write a more formalized, rehashed version of Ryan's answer
 
vzn
3:36 AM
ryans answer while insightful is not nec definitive, its a reasonable sketch. a more definitive answer would be something that was published somewhere.
it appears you interned at MIT? are you still working with aaronson or plan to work with him further? it looks like youve taken a few grad classes as an undergrad?
see that you have a paper-in-progress on minecraft. any opinion on the MS acquisition? am planning on blogging about it sometime....
see some FPGA work too, have chatted with WL on that topic here.
 
mdx
4:13 AM
The portfolio is incredibly out of date. I plan to redo the entire site soon. Some stuff got reformulated, added, or scrapped for various reasons (life, interests, etc).
Aaronson and I still keep in touch. My plans for the future are a little unclear atm. I drafted out the Minecraft paper, but 1.6 got released (followed by 1.7 and 1.8). I would have to redo the proof that generalized Minecraft is not Turing complete to account for slime blocks (huge impact). I wasn't too happy about the acquisition. MS needs to realize how thin of ice they are on.
 
vzn
4:38 AM
are you still in school? did you graduate? yeah the acquisition is surely quite controversial to say the least.
havent heard of any TCS analysis of minecraft at the moment other than yours. have you? your analysis of a fixed version of minecraft seems worthwhile & notes at the end could fairly point out that it has changed in later versions.
there are a few game analysis questions on Computer Science and Theoretical Computer Science, some highly rated... have tried to collect them in the blog
MS is going thru tough gyrations at moment. just announced closure of 13yr old TCS lab, Silicon Valley. many elite researchers brusquely laid off.
 
mdx
Still an undergrad. I have to finish up the prereqs I skipped. The graduate level courses unfortunately don't get to count towards my degree. I've considered moving on in life instead of finishing undergrad. There are some personal stuff getting in the way of maintaining my "concentration".
 
vzn
holy cow re "not counting". thats terrible to hear. took a grad class as undergrad myself & got credit toward graduation for it.
 
mdx
I haven't looked into it in a while (over a year), but the Minecraft result would be a bit different than the other game analysis results. Most of the other ones I've seen focus on a well known kind of puzzle or create one based on the setting of some of the games. I considered whether universal computation (generalized scale) is possible using the game mechanics.
 
vzn
there is a long history of research very similar to what you mention, incl undecidability results found in games.
one of the earlier ones is Conway proving Game of Life is Turing Complete.
~1970s
 
mdx
Ah, by games, I was focusing on video games :)
Conways is too old to be considered a real game, haha
 
vzn
4:48 AM
Game of Life is rather close to a video game :)
agreed it is primitive in comparison to "modern" video games.
 
mdx
Have you seen http://www.youtube.com/watch?v=3UnB1fomvAw ?
Not exactly related, but interesting none the less.
 
vzn
but there is serious academic analysis of "modern" (video) games. not sizeable but its scattered out there.
TM in pokemon? cute :)
in cyberspace everything is related lol
btw many papers consider "generalized scale" games.
 
mdx
I'm trying to separate out situations like showing that Pac-Man is NP-hard because finding optimal routes is (TSP). There are other examples such as the "gadgets" in the Pokemon NP-hardness proof or pressure plate/switch puzzles. The way that the scenarios are generalized are a little bit artificially tuned to the complexity analysis imo
In other words, generalized instances of puzzles present in games != computational power given to the user through game mechanics
But yeah, it's been a long long time since I've looked around for similar research focusing on modern games. I'm sure there's more out there by now
 
vzn
5:04 AM
yes the step of generalizing the game is glossed over sometimes. there is more than one way to generalize the game.
maybe there are something like "game complexity classes" waiting to be mapped out.
you clearly have some great theoretical talent & drive & think it could be some loss if you leave academia.
 
mdx
5:18 AM
Thanks for the compliment :)
I haven't made decisions yet, so maybe I'll stick with it. Perhaps I'll find other opportunities to apply that kind of thinking in other places, though.
I see from your blog that you are a software engineer by trade. What got you interested in cc?
 
 
4 hours later…
9:49 AM
@mdx You can post an answer with a summary and a link to the original and mark it as "Community Wiki" (so you won't get rep).
That is, if you want to clearly abstain from taking credit. It's absolutely fair if you post your summary in "normal" mode.
You can also post the answer you were preparing independently.
Your choice.
By the way, consider finishing your degree; even if the degree itself is only a sheet of paper it can open doors.
 
 
5 hours later…
vzn
2:58 PM
mdx, what is "cc"? did you mean cs?
lol "degree only a sheet of paper"
if only
could have saved so much trouble with just a photocopier :p
yes mdx no good education goes to waste & even partially completed schoolwork has some value although in life there are various key "theshholds/ milestones"....
yes there are many examples of sometimes eminent leaders as college dropouts esp in the tech field, but they are obviously exceptional in many ways....
latest college dropout billionaire: zuckerberg. earlier ones: gates, jobs.
 
@vzn "cc" could be computational complexity
 
vzn
ok
oracle ceo ellison also college dropout.
here ofc in these cases "correlation≠causation"
 

« first day (930 days earlier)