Datasets:
cairocode
/
MSP_POD_NLP

Dataset card Data Studio Files
xet
 Community
2
text
stringlengths
0
353
EmoClass
stringclasses
10 values
EmoAct
float64
1
7
EmoVal
float64
1
7
EmoDom
float64
1
7
SpkrID
stringlengths
1
7
Gender
stringclasses
3 values
it's interesting because when usually people talk about consensus-
N
4.4
4.2
4.4
1656
Male
they start with distributed systems and then they go into blockchain. but-
N
3.8
4
4.4
1656
Male
... starting with the blockchain part of it and then realizing that this field of consensus research has been there for over 20, 30, 40 years.
N
5
3.6
4.6
1656
Male
so a lot of cynthia dwork's work. and this was a very, very profound beginning of consensus in stronger adversarial models.
N
4.8
4.4
3.6
1654
Male
... nodes would just go offline and online.
N
4.6
3.8
4.2
1653
Female
so how did this field progress? and how did this capture people's interests over so many decades?
N
3.8
4.2
4
1653
Female
... really impactful way has brought out technologies like zero-knowledge-
N
4
3.8
4
1653
Female
... proofs. has brought research that people would do in probability-
X
4
5
5.2
1653
Female
... or distributed systems back into the mainstream as something that's really-
N
4.6
4.2
4.6
1656
Male
... really important. i think people back then researched it because it was definitely an interesting topic of study-
N
4.6
4.6
4.4
1654
Male
but i definitely think blockchains have had a huge impact in reinvigorating this kind of research.
N
3.8
4
4.4
1654
Male
the way i got into research was into consensus was through blockchains-
H
4.8
3.8
4.2
1653
Female
so i started looking at different proof-of-work and proof-of-stake based blockchain protocols.
N
4.4
4.4
4
1653
Female
so a lot of people might have heard of definity, of tendermint-
N
4.4
4.4
4.8
1654
Male
... ethereum's new protocols, bitcoin. i would say the focus for mechanism labs has so far been on proof-of-stake. or proof-of-stake, proof-of-work-
N
4
4.6
4.6
1656
Male
started out with proof-of-work. and one thing we saw as a huge problem was this huge energy and efficiency-
N
4.8
3.6
4.6
1656
Male
i personally don't know if mining is long-term sustainable. and when you, in general, think of building a technology that is scalable across generations, across people, across countries-
X
4
3.6
4
1656
Male
what are other externalities of such a technology?
N
4.4
4
3.8
1656
Male
i think that mining, in and of itself-
N
4.4
4.4
4.6
1656
Male
... is not necessarily any more efficient than today's economic systems or financial systems-
A
4.6
4
3.4
1656
Male
... anything, it's probably going to make all the systems that we've built and optimized over the years more inefficient.
N
3.8
3.2
4.2
1654
Male
i think we need to be building something that brings out long-term sustainability and probably is way more efficient than what we have today.
N
4.4
4
4.4
1656
Male
... proof-of-stake. and i think proof-of-stake was put forward as one of the very earlier models that-
N
4.8
4.4
4
1654
Male
... were long-term energy efficient. recently-
N
4
4.4
4.6
1653
Female
these are all newer ideas and newer concepts. so that's why mechanism labs chose to focus on proof-of-stake.
N
4.6
3.8
4.6
1656
Male
yeah. so i guess one way of thinking about any of these consensus mechanisms is when you really say consensus-
N
3.8
3.8
3.8
1656
Male
... which is how do you prevent people from creating multiple identities?
N
3.8
3.4
4.2
1656
Male
and the second part is, how do you actually come to agreement, but wean the people who have been chosen to come to agreement-
N
4.2
3.8
3.8
1656
Male
... and when you say proof-of-stake or proof-of-work or proof of elapsed time or any of these other things-
N
4.4
4
5.2
1653
Female
... you're essentially trying to have some sort of civil control mechanism. with proof-of-work that is by burning energy. with proof-of-stake-
X
4.2
4.6
4.8
1653
Female
and i think each of these is just you're putting down a different resource to then select different people who would be part of the group that comes to-
N
4.4
4.6
4.6
1653
Female
yeah. yeah. the actual part where people come to agreement is still-
N
4.8
3.6
4
1656
Male
you have no concept of probabilistic finality.
N
3
3.4
3.2
1654
Male
... goal is for all honest nodes to come to agreement on the same view of the network, versus in blockchains-
O
4
3.4
3.8
1656
Male
... you have this concept of all the nodes may, with a high probability, come to a common agreement of what the network looks like.
N
5.2
3.2
4.6
1656
Male
yeah. so these are the traditional ways of coming to consensus. what bitcoin uses-
N
4
4.6
3.8
1656
Male
... and each of these, the reason i call them probabilistic consensus is because of china and-
N
5
4.4
4.8
1653
Female
... is secretly mining the longest proof-of-work chain-
N
3.8
3.4
4.4
1656
Male
the moment that they come out with the longest chain-
N
4.8
3.8
4.4
1656
Male
... or that gets discarded, versus in traditional consensus-
N
4.8
3.8
4.2
1653
Female
yeah. and i think the other interesting part, though-
N
4.2
4.2
4.2
1653
Female
it obviously depends on what sort of pairing any blockchain consensus uses. like it-
N
3.8
4
4.2
1653
Female
... depends on what rule they use to come to agreement. but-
N
3.8
3.6
4.2
1653
Female
... with bitcoin, you have this ability for anyone in the world to be part of the bitcoin mining community to join and suddenly join the-
N
4.4
4.2
4.6
1656
Male
and this idea of permission-less is something that blockchains have enabled and-
N
4.2
4.2
3.8
1653
Female
and in order to have these additional properties, you have to trade off some things-
N
4.6
3.8
4.6
1656
Male
but there's also got to be some inefficiency that's introduced by having node churn by [inaudible 00:09:31] who is constantly dynamic and-
N
4
4
4.4
1653
Female
... decentralized systems are way less efficient than centralized systems.
N
3.6
4
3.8
1653
Female
... every block, you have a new committee. churn definitely has its disadvantages.
N
4
3.8
5
1653
Female
but one huge advantage that i see a blockchain as a technology is that it's there. it-
N
5.2
3.8
3.8
1653
Female
so you don't necessarily need a blockchain to be the most efficient application.
N
4.4
4.4
4.2
1656
Male
and in a sense, having the ability to manage a huge number of people who want to join and leave-
N
4.4
4.2
4.2
1653
Female
i think is really impactful because it provides a large number of people to keep the system up.
N
4
4.6
4
1653
Female
yeah, i definitely think even in building out the dream of block-
N
4.8
4
4.4
1653
Female
blockchains, we can build a more optimized-
N
3.6
4
4.2
1653
Female
exactly, and also, i don't know if bitcoin would be long term sustainable and would be a certain fallback.
N
4.4
4.4
3.8
1656
Male
especially if mining itself is not something which everyone can do and is-
N
4.6
3.4
3.4
1653
Female
proof of stake seems like a really interesting concept.
N
3.8
4.4
3.8
1656
Male
so could you define proof of stake and differentiate it to proof of work?
N
3.6
4.2
4.4
1654
Male
to be part of the consensus of that network.
N
4
3.6
3.8
1654
Male
you're incentivized to act in an honest manner in order to have the value of your coins appreciate or-
N
3.8
4
4.8
1654
Male
if you act dishonestly, the value of your coins go down. versus in proof of work-
N
3.8
4
4.2
1654
Male
[inaudible 00:01:17] or at least that was satoshi's idea, but-
N
4
4.2
4.2
1654
Male
yeah, i guess the broad categorization that a lot of people use is chain based or bft based.
N
4
4.2
4.4
1656
Male
pbft stands for practical byzantine fault tolerance.
N
3.8
4.4
4
1654
Male
it's almost like a curiosity that had these kind of very niche.
X
4.2
5
4.6
1653
Female
applications where it might have been extremely valuable and it's only just now that they've suddenly found this highly financialized purpose that achieved all of this renown.
N
4.2
4.6
4.6
1656
Male
i would definitely say that was quite the case.
N
4.4
4.2
3.6
1654
Male
just like if you look at other cryptography or other research that was going on, zero knowledge proof-
N
4.2
3.6
5
1656
Male
if i'm not mistaken, and it's only very recently that the applications of it are coming to light.
N
4
4
4.6
1656
Male
was the different types of proof of stake protocols.
N
2.8
3.8
3.6
1656
Male
yeah.that from memory, i think that was the first proof of stake consensus protocol.
N
4.4
4.6
5
1656
Male
yeah, so we definitely did not include them in our analysis.
N
3.8
3.8
3.2
1653
Female
it seemed a lot more primitive and there definitely was a larger possibility of attacks on those early protocols. like the idea of stake grinding attacks where a minor could-
N
4
3.2
4.4
1656
Male
basically go back in time and try different possibilities of what block would give him the chance to be the block creator for the next round.
N
3.4
4.4
4.2
1654
Male
seemed to be a huge attack vector on these early protocols.
N
3.4
3.6
4
1653
Female
and that was a huge problem with these early protocols.
O
3.4
3.6
4.4
1656
Male
and so this is what was happening in 2014 then.
N
3.6
4.2
3.8
1654
Male
that's one of the reasons we didn't choose to include them in the analysis.
N
4
3.6
4
1653
Female
and they gave rise to different protocols that we see today.
N
3.8
4
3.8
1657
Male
like affinity, and tendermint, and thunderella, and all these really cool projects-
H
4.4
5
4.6
1653
Female
but there were these really obvious and blaring attacks that definitely didn't seem any more sustainable than bitcoin. if-
S
4.2
3
4.4
1653
Female
casper, like ethereums, caspers, thunderella, and these are definitely some examples of the newer generation of protocols.
N
4
4.6
4.2
1653
Female
it's really cool. they use threshold relay and they use this decentralized way of creating randomness.
H
4.4
4.8
4.4
1653
Female
and then you have like tendermint, which basically doesn't use randomness at all.
N
4.6
4.6
4.4
1657
Male
[inaudible 00:03:37] all these use some really clever like randomist generation to prevent stake grinding attacks.
N
4.2
3.8
3.6
1653
Female
[inaudible 00:03:44] cannot cause finality to break in the protocol, but-
N
4.2
3.8
4
1653
Female
for example, in tendermint you can have at most a third being malicious.
N
5.4
3.6
3.8
1653
Female
if you have less than a third of the stakeholders or a third of the voting power malicious.
C
4.6
4
4.8
1657
Male
or malicious, then even those 10%.
X
4
3.2
4.6
1657
Male
if you threaten people saying that i have 10% of-
N
4.2
3.6
4.6
1653
Female
mining power, i will not mine on top of your chain.
N
4.2
3.8
4.4
1653
Female
that in and of itself is incentive enough to prevent someone from mining on top of a chain that includes arthur's block. so no one will want to include arthur's-
N
4.4
3.2
4.4
1653
Female
it's a kind of pothole to break your ankle as you go down this road-
N
4.2
4.6
4
1653
Female
definitely. and i think what makes it harder is the fact that you have layers upon layers of complexity. you have-
X
4.4
4
4.2
1653
Female
bringing in this blockchain component, which requires civil control mechanisms, in order to have this trend component. and when you compose all of these different things together, it-
N
4.2
3.8
3.8
1653
Female
it's not just the security of individual components, which is important.
N
4.4
3.6
3.6
1653
Female
i definitely see proof of stake as a way of both combining this reward distribution mechanism and the incentives around that.
N
4.8
3.6
3.6
1653
Female
that's kind of separated out where the coin itself is different.
N
3.6
4.6
4.2
1653
Female
that's been quite challenging because, one, it's hard to get this combination of both economic security along with cryptographic and distributed system security.
N
3.4
3.8
4.2
1653
Female