How to Fix Doge (For Elon Musk)

hi this is charles hoskinson broadcasting live from warm sunny colorado always warm always sunny sometimes colorado it's good to be back in the office and what i had a lovely time in miami i was noisy and lots of people running around and way too much skin and i tried to spend most of my time reading a book from alex alexis burgess on truth and i never had the opportunity to do that but i did spend about thirty dollars for a coconut drink that had rum in it and i was very happy with my coconut rum okay so some housekeeping picked up a lovely new book from samuel mimram it's a program and proof and it's actually a book that kind of introduces the curry howard correspondents and it teaches you how to write software in both o'camel but oddly acta usually you see that paired with and i can't wait to look at it and send samuel a loving note okay so the purpose of this video is to if elon is actually serious about rebuilding doge with his elite ninja engineers over at tesla which i have no doubt he has great programmers there here's how you do it so let me share my screen because i got some great papers here and it's gonna be fun you guys ready okay let's start with the engine right now doge operates on a pretty shitty nakamoto consensus style with script engine because i think it's a fork of litecoin so let's tear that out put something in prism isn't probably the best and fastest proof-of-work protocol on market right now and if you put it into a bitcoin style system you can get about 10 000 transactions per second it's written by an all-star team promote ms swanath over at urbana-champaign you got people at cmu stanford university of washington and i we really like david as well david she's a great guy so prism is a phenomenal protocol we've implemented it ourselves in scala and highly recommend looking through it's a beautiful paper that talks about trade-off okay so it's not good enough just build an engine the whole point of litecoin with being based on script was to be asic resistant right well then you need an asic resistant algorithm and actually this one's really cool it's called hash core and i'll just read off some of the abstract the past five years rewards associated with mining proof of work blockchains increase substantially as a result miners are heavily incentivized to design and utilize asics that can compute hashes far more efficiently than existing general purpose hardware currently it is difficult for most users to purchase and operate asics so the point of this paper is basically to come up with a really cool way of getting around them so in order to achieve this we present hashcore a proof-of-work function composed of widgets generated pseudorandomly at runtime that each execute a sequence of general purpose processor instructions designed to stress the computational resources and so it's really hard to actually build an asic against this and what i like about it is made by electrical engineers so you can also look at some beautiful theory and this is from dan bonnet who's probably one of the top cryptographers in the world over at stanford and it's called balloon hashing a memory hard function providing provable protection against sequential attacks so this is another primitive that one can use while discussing the notion of asic resistance are we done yet no we got this concept of difficulty control in blockchain systems this is a paper written by some iohk alums dimitri meshkov and alex cherpinoy and basically it talks a bit about how to have a much better way of doing difficulty retargeting because if you're going to a new type of proof of work with a completely new system that shards things and runs at 10 000 tps you need to think about difficulty targeting okay are we done yet no we need to change the block structure so in the header put coded merkle trees this helps you a lot especially when paired with here we go nippa pals non-interactive proofs of proof of work these two things together in a proof-of-work system give you a beautiful path for extremely efficient like clients as well as a great primitive for side chains then you have mining and logarithmic space this is a new one actually real sexy really exciting paper by aguelos and nico leonardis and dionysus xindros and this is basically saying you don't need the blockchain to mine the blockchain imagine that prism 10 000 tps don't need no blockchain you got beautiful side chains no data availability issues your light clients are there and then you have a new way of doing difficulty retargeting instead of the old crappy stuff that they do in narcomoto consensus and you got two very good paths to pursue for asic resistance to replace script then we have to put a good scripting language in because doge is not programmable and let's get some smart contracts for utxo model and not going down the crazy road we went down with cardano which would be too much for your guys i think what the ergo guys have done is absolutely phenomenal and there's a lovely paper on ergo script non-interactive zero knowledge proofs supported with erco script and there's all these great little papers here about sigma protocols and so forth and there's actual code running live right now on ergo you guys would love it now another thing the network stack of doge is old and ossified so go ahead and read this beautiful paper it's an sok paper sok papers are roll-ups of all of the latest and greatest in knowledge and it was written by avi zahar who's a wonderful guy he's the guy that did spectre and phantom and ghost and all those other directed acyclic graph protocols to accelerate proof of work and then you got some dfinity guys you got some people in switzerland you got some people in vienna and this really takes you through the latest and greatest of all the different network stacks on market trade-offs and various things to know so you guys can figure out your head from your ass on how to handle the network side and make it better because this boy right here is going to put some major strain and you can't just throw secure academia on it are we done yet no come on let's keep going let's take a look what nist is doing recommendations for stateful hash based signatures why because doge should be quantum resistant right and you got some great things like mcauley's signature scheme lms or extended merkle signature schemes or you got hss or actually the new multi-tree xmss you have a blockchain you can put a stateful based hash signature scheme it's becoming a nist standard and that means that you are quantum resistant and then you have blake three because why be with an old shitty hash algorithm this is the age of doge so let's go ahead and put a better hash algorithm in and the best guy around for hash algorithms is zuko and he created blake three makes mdm fi it makes md5 and sha2 and shaw three look a sack of so highly recommend this so there you go and just a few easy steps one two three four five six seven eight that doesn't count eight nine ten eleven you guys are off to the horses running and making doge better for everyone inside this space 10 000 tps you get some quantum resistance you'll be able to figure out a network stack that'll meet your needs and satisfy your urges you'll have programmability look at that you can finally do smart contracts on doge and have all those doge related applications you don't need the blockchain to mine how about that you get side chains you get some cool stuff for like clients and then of course you'll have a much more reasonable difficulty system retargeting system and you're asic resistant this is a fun paper so i hope this is helpful for the doge community and for elin and i wish him the best of luck in rebuilding that cryptocurrency with his super elite engineers which i have no doubt he has as he's one of the world's richest men thank you truly for entering our space and thank you truly for telling everybody that you're gonna take custody of that cryptocurrency and make sure it gets where it needs to go and i look forward to seeing the great innovations you guys come out with in a year or two cheers