00:00welcome to modular March I'm handy this
00:02is Robbie and we are the rollup modular
00:04March is an educational experience about
00:41here today with John jcole not the
00:45rapper from hyperlink if you watch our
00:47recent interview we had a really short
00:48conversation about modular expansion and
00:51today we're gonna expand upon that so
00:53John welcome in um and would love to
00:56learn more from you today and hear about
01:00yeah no Andy it's good to be back always
01:02fun to be on here with you boys whether
01:04it's you know when we're recording in
01:06person like we did in Denver or you know
01:09another more like Studio episode like we
01:11did before I'll uh I'll spare folks you
01:16know a broad introduction and like life
01:18story if people want that definitely
01:19they can check out the first episode we
01:21did uh we had a bit too much fun on that
01:23one um so I'm here from hyperlane in
01:27like just a little bit of context what
01:30we're doing with hyperlane is we are
01:32building the world's first uh truly open
01:35framework for interoperability something
01:38that should let anyone connect any chain
01:40and do it on their terms kind of what I
01:44is um yes in 2024 there are ways for you
01:49to connect your chain with others but it
01:52involves generally like this offchain
01:55back room conversational BD type process
01:58you need to go and prove to one of these
02:00interoperability providers they kind of
02:02worthy of their time investment and
02:05despite your best efforts the moment you
02:07get outside of the IBC ecosystem like
02:10you can do things yourself you can
02:12maintain you can't do them on your terms
02:13you can't do it uh well maintaining your
02:16sovereignty until hyperlane comes in and
02:18so what we're doing is we're taking what
02:20was an offchain and kind of opaque you
02:23know BD process and we are putting it on
02:26chain we're making it transparent we're
02:28anyone uh do it so to your first
02:31question like what does modular
02:34expansion really mean to us well you
02:37know the reason we started kind of using
02:39that phrase and throwing it around
02:42because we've had what I consider to be
02:45like the big bang of the modular
02:47ecosystem uh somewhere in Q4 last year
02:50call it right around beginning of
02:52November that was the culmination of a
02:54bunch of events Celestia's launch was
02:57the headliner it gave us the first
02:59alternative da layer and so really open
03:01the gate for it but it wasn't just that
03:03it was a few other things all coming
03:05together at the same time notably right
03:08the continued like maturation of the
03:11rollup Frameworks right so like the
03:12amazing work by people at arbitrum and
03:15optimism with the rollup
03:18Frameworks the great work by the rollup
03:20providers like Caldera and conduit and
03:23ALT layer and others all of these when
03:26put together made it really really
03:29really easy for the first time to create
03:31your own chain your your own rollup so
03:33much so that like it could actually be
03:35done in you know hours to a couple
03:38days and soon will probably be minutes
03:42versus what was a multi-month process
03:44beforehand and so start seeing okay this
03:47big Bang's happening or we're getting an
03:49increasing number of
03:51rollups now this is moving towards the
03:54expansion phase you're starting to see
03:55these coming out on a regular Cadence
03:58but what we've believed in hyper
04:01what we've argued is that like it
04:02doesn't really make sense to talk about
04:03the modular sa without talking about
04:05interoperability you have all these
04:07Trends and all they're doing is like
04:09driving down the cost of creating the
04:12rollup well that's great that means that
04:16more and more rollups are going to be
04:17created but how do they connect to each
04:20other how do they connect to existing
04:22chains how do they transfer value like
04:26you think of right A lot of people I'm
04:28one of them like the anal of like
04:30blockchains As Cities as population
04:34like how many people want to move to a
04:36new city that's in the middle of the
04:38desert and has no airport no roads you
04:41can't connect you know like there are
04:42some people who would want that who
04:43completely want to live like Off the
04:45Grid generally speaking you want to have
04:48Commerce and trade routes and you want
04:50to have a way to go between that City
04:52and the established one that's how
04:54cities can grow much more much more
04:56quickly and so hyperlane by this virtue
05:00of it being a completely open framework
05:02that anyone can use to connect any chain
05:05now serves this role of being the
05:06connective tissue and the thing that
05:09really accelerates the modular expansion
05:12and so that's um kind of how we view it
05:15I'm happy to get into like much more
05:16detail about what that means in practice
05:18how we help uh bring this about how we
05:20work with different teams and you know
05:23anything you or the audience might want
05:26know for sure so when it comes to kind
05:29of connecting uh these execution
05:31environments these rollups together is
05:34there any process that if I'm a new
05:36rollup that is launching is there any
05:38process that I have to go through in
05:39order to be deemed um like uh ready or
05:43appropriate to connect to hyperline such
05:46that other connections won't be affected
05:49if I was a uh malicious rollup or if I
05:54down great great question um the like
05:59most honest clearest answer there is not
06:01no because hyperlane is a totally open
06:05framework anyone can use it to bring it
06:07to any chain and now that chain is um
06:11kind of eligible think of it as now
06:12having the correct Parts the right so
06:15you know uh plugs and sockets to connect
06:18with any other chain that has
06:21hyperlane but does that mean that like
06:25hey if I put hyperlane on my chain Am I
06:27by default connect it to Every Chain in
06:28the world that has hyperlane no it does
06:32not so in hyperlane like an integration
06:36entails three moving Parts uh and let's
06:41cover them like at a really high level
06:43one hand we have the mailbox think of
06:44this as like the base contract that
06:47people interact with this is the things
06:48that sends and receives messages between
06:53um it's completely on chain and so far
06:56as like the different p uh parts of the
07:00part is a security module and hyperlane
07:03instead of saying like Hey we're the
07:05smartest people in the world we know how
07:07to secure this thing we figured the
07:10crypto industry is full to the brim
07:14right it's teaming with innovators on
07:17the security front so much so that like
07:19the state-of-the-art of this year is
07:22going to be junk by next year right uh
07:25for myself you know I've been around
07:28since 17 and like you could really see
07:30how every year you know the the
07:33state-ofthe-art kind of obsoletes
07:35everything that's come before it and so
07:37it was imperative to design a system
07:39that really could absorb anything any
07:41type of security mode the system can
07:44adapt to so a hyperlane security module
07:46really what it is it's a basic wrapper
07:49on anything that provides either a state
07:51at station think like a validator system
07:54system um or a state proof think like
07:58you know some light line verific
08:01magic anything of that sort so we need
08:04at least one security module to be
08:06operating on the Chain if we're going to
08:08have an integration so that was the
08:10first two parts mailbox second part
08:12security module third part uh is the
08:16relayer it's basically like an offchain
08:19bot it looks at the different chains
08:21that you want to connect so let's say
08:23like your chain and Robbie's chain and
08:26it just needs to look at the mailbox it
08:29sees Andy is trying to send something
08:31from his chain to Robbie's I see Andy's
08:33a nice guy he's paying this
08:35fee cool I then look at the security
08:39module that is uh operating on Andy's
08:42chain and I see that it's signed off it
08:44provided these signatures I take those
08:46signatures to the mailbox on Robbie's
08:48chain and now mailbox processes that
08:50message checks if the signatures are
08:52valid and So within like if you have
08:56these three parts operating that means
08:58you have a hyper L connection
09:00established but in practice what happens
09:02is with those three things there any
09:05developer can now use these Pathways but
09:09when they do that what they're saying is
09:11like let's say someone comes to to your
09:14Andy they now can connect with any other
09:16chain that has hyperlane but they need
09:19to Define and they need to say hey I
09:21want to accept something from ethereum I
09:24see that there's these three or four
09:26different security modules running on
09:28ethereum I like module number B the best
09:32it's my favorite one it seems to me the
09:34most secure maybe that's the one that
09:36runs like with some ZK magic and that is
09:39the module that I'll accept for my
09:42applications and so in that sense it's
09:45very much like an optin system right so
09:49no one on Andy's chain yet received
09:51something from hyperlane just because we
09:53put it on there uh no one can deny us
09:56from receiving something but once it's
09:59there now as a developer I can choose I
10:01can say hey I want to receive messages
10:03on Andy's chain I want those messages to
10:06come from ethereum and I want them to
10:07come through like that module and so in
10:12that sense like someone would basically
10:16have to um create kind of create like a
10:20malicious module and then fool people
10:24it and so that of course can happen
10:26right like bad people do you know there
10:29scams everywhere there are bad people
10:30doing bad things um but that's the I
10:35guess the set of trade-offs you get with
10:36a totally permissionless system so we
10:38built a system that's very
10:39permissionless we create it in with this
10:42very much like an opin type of design
10:44but still still people can opt in to
10:47something that is uh fraudulent like
10:50there was one time in my uh you know
10:53early Beginnings in crypto I opted in
10:55with my money into some wonderful coins
10:58that I thought were great uh like deep
11:00brain chain great project from 2017
11:05and I may have opted in something that
11:07wasn't very real uh but you know I I
11:10certainly thought it was very real I was
11:12totally into deep brain chain I remember
11:19yeah I think sure it's still going I'm
11:22sure they're still working
11:24hard that's the beauty of blockchains
11:26people don't understand is that they
11:28just run in perpetuity as long as some
11:30some entity or entities are validating
11:32the network or mining it so I'm going to
11:34pick and a theyve been building AI ever
11:38man G to pick and PR a bit about kind of
11:41like this wrapping process so basically
11:43what you're explaining is that you're
11:44almost you're almost uh like an API
11:47rapper for uh a lot of these projects do
11:50you feel that like there's some sort of
11:52um uh problematic scenarios which can
11:56come from kind of the that wrapping
11:58process of these of these different
12:01projects totally yeah now um I think you
12:05know right like from our perspective We
12:07Believe very much in like transparency
12:10and and uh in openness maybe it's like
12:12the direct nature of like the Israeli
12:15and me but it's it's much easier to just
12:18like start off and talk about all the
12:19things that could go wrong right and
12:21like just being very upfront about so in
12:25this world anyone can create can create
12:27a module right and and we hope that the
12:30people who create a module do it on the
12:32up and up and that what they're doing is
12:34like they're saying hey either I created
12:37a certain security innovation or I have
12:41observed an open- source security
12:43innovation and I want to rrap it and
12:46so um based on that we hope that
12:50everyone will do the right thing and
12:52we'll just actually wrap through
12:55Innovations as security modules but
12:58someone else could say hey look at me I
13:01follow the instructions I have created a
13:03module that you know I call I'll call it
13:06the um wonderful ZK module and so
13:12like they could totally like wrap
13:15something malicious right where they say
13:17that it's one thing but in practice it
13:19is absolutely not that thing they could
13:22advertise for it they could do
13:24everything in their power to make it
13:26seem legitimate maybe even operate it
13:30for a long while and kind of you know
13:33create a sort of like long con and then
13:36one day after they've had enough uh you
13:39know hardworking honest developers kind
13:43module they could rug uh in the same way
13:47that like you know a project can like a
13:50defy project can say you know uh run
13:53legitimately for a really long time and
13:55then just like rug all of its depositors
13:58or something like that
13:59and so it is totally possible and like
14:02the things that we can do to combat that
14:05are more again keeping everything open
14:08source like never never ever allowing uh
14:13for a module to like receive some type
14:16of stamp of approval unless like all the
14:18contracts are verified unless everything
14:19is open source unless everything is like
14:22totally visible um I would say actually
14:25like I'm less concerned at this point
14:28with being malicious then I am concerned
14:30with just honest to God good mistakes
14:33you know like people who are
14:34well-intentioned and then still things
14:37go wrong right like if you think about
14:40most of the security incidents in crypto
14:42especially in our sector almost all of
14:45them are actually like exploits that are
14:48born out of a CO you know a mistake in
14:52code and so those are much harder to to
14:56prevent right like if someone's doing
14:57something nefarious like
14:59you take an eye to it you might be able
15:01to spot it and it would take a really
15:03sophisticated like long con to to fool
15:07you know talented developers that code
15:09that isn't malicious sorry that code
15:13isn't what's much harder to catch is
15:17like just a tiny tiny mistake somewhere
15:20that in the right conditions can be used
15:23to like upend the entire thing and
15:25that's always like the danger of again a
15:29totally open and permissionless system
15:31it's why like actually if you think of
15:34the world of like Computing in general
15:36you know uh on PCS everything was like
15:40totally open you could install whatever
15:42you wanted but there was you know like I
15:46guess I'm a little you know I'm old now
15:48but uh all my friends I was like The
15:52Tech Guy or you know like the tech nerd
15:54and people would want me to come install
15:56stuff for them or like when they were
15:57downloading they would ask me my buddy
15:59like where should we download from
16:00Because the Internet was like a very you
16:03know dangerous place you didn't want to
16:07machine uh then with the Advent of
16:11smartphones and the App Store you
16:13finally got like a much more curated
16:15version and so it was sort of open like
16:18anyone could create something and could
16:20petition to be on the store and then
16:22could go through a process and yes they
16:23could get revoked at any moment in time
16:25sometimes without any notice about why
16:28but it actually like I think Advanced
16:30Computing because it opened it up to a
16:33much like less sophisticated
16:37crowd um and right like people didn't
16:40have to worry like you rarely worry that
16:41something you download from the App
16:43Store will be malicious and
16:47so that had to be like a curation
16:49experience from like apple a massive
16:53company that even at the time when it
16:55wasn't you know one of the 20 largest
16:57companies in the world and wasn't one of
16:58the larg just two like it is today they
17:01still had the resources to enable that
17:04but with a crypto system how can we use
17:06like the incentives uh that you know
17:09crypto crypto systems afford US to
17:12create a much more open version of this
17:15curation such that like maybe you have
17:17incentives towards like call them like
17:20different little Watchdogs which is like
17:23trying to decide which are the clean
17:25modules like how do we you know keep an
17:27eye on them and expose and Elevate the
17:30good and clean ones and completely
17:32distance ourselves you know meaning as a
17:35community from the bad ones either
17:37because they're you know either they're
17:38bad because they're genuinely malicious
17:40or they're bad because like there's just
17:42an error there that no one's caught like
17:44how can you run a sort of like immuni or
17:47like cod Arina process on these
17:49different modules is there a
17:53was just real quick I was trying to to
17:56get into that because you guys are
17:57pushing like permissionless very hard um
18:00which is great and I think that's very
18:02important for the space um but if you if
18:05you kind of are missing some of the
18:06security components you have to either
18:09like isolate the risk somehow of of of
18:11these uh different chains as you kind of
18:13said that you've done or basically what
18:15you have to stop is you have to stop
18:17somebody from creating a chain
18:18integrating the Hyperion SDK and just
18:20like basically mincing a bunch of assets
18:23um you know from just a random chain um
18:28yeah because I mean there's no yeah now
18:30the good thing that that actually like
18:33isn't a huge concern because again like
18:35people have to kind of opt in to receive
18:37a connection from that chain and accept
18:39those like and so that that part like in
18:42and of itself is not very concerning the
18:44more concerning one is like the two I
18:46mentioned before it's either like a
18:48malicious you know person creating a
18:50module that looks you know genuine and
18:53then like letting you use it and then
18:55one day just like pulling it or a
18:57totally wellend mentioned good you know
19:00good actor that just misses something
19:02right like sucks but that's probably
19:05going to happen uh at some point right
19:07like it's happened to some of the best
19:10like uh you just have to have like
19:13security practices uh at all fronts and
19:16like no team can get you can never like
19:19um have like hubis on your team such
19:22that like you know maybe there's someone
19:24who has an elevated status in a team
19:27it's like oh well we're not going to
19:28review their stuff because like they're
19:29super they're so good like they're a
19:31security expert they would never make a
19:33mistake like everybody makes a
19:34mistake so they are like certain best
19:38practices you can do in developing the
19:40software to mitigate that and then I
19:42think as a you know as a secondary and
19:44perhaps even more important measure like
19:46how do you tap in to the structure of
19:49these open networks to the advantage of
19:52everyone who uses them right like one
19:54way is to let anyone create the modules
19:56but another is like let anyone audit
19:57them let anyone like kind of Express an
20:00opinion about which ones should rise to
20:02the top and which ones should kind of
20:05flounder so so you you mentioned how
20:07it's not so much about the chain uh
20:10because people opt into which chains
20:11they want to connect to it's more about
20:13the security modules and right now it's
20:16permissionless to create these security
20:18modules are you doing any and you
20:21mentioned that like anyone can audit
20:22them which is kind of like the
20:23validation mechanism to check double
20:25check if there's any vulnerabilities
20:28with any of those modules is there an
20:30incentive mechanism or generally how do
20:34you how do you think about the potential
20:37vulnerability of a either malicious or
20:41buggy security module like what is the
20:43current practice to mitigate that now
20:47and if there is any practice I'm curious
20:49if there's an incentive mechanism that
20:52incentivize secure modules so there are
20:56two processes right now which are
20:59uh the first which I mentioned is
21:03internally multiple eyes on it like
21:07anything that is going to touch funds
21:09anything that's going to like serve as
21:11the substrate for different modules goes
21:13through multiple reviews and multiple
21:17reviewers um and that
21:20includes yes because um right like
21:24before something is kind of merged at
21:26least like the core team will have
21:28multiple I on it now again we can't stop
21:31you from like doing something in your
21:33own repo but like if it's going to get
21:36merged into like the core repo right by
21:40now it's not just like our uh core
21:42development team that runs that there's
21:44a bunch of different
21:45contributors um to that repo but still
21:48like try to get eyes
21:53sources and then of course like uh
21:57audits on some of the more prom ones and
21:59the second thing is kind of you
22:01mentioned incentives like having open
22:03bounties um through systems like imuni
22:07but I don't think it can end there I
22:08think there has to be uh something more
22:11so now like us in conjunction with other
22:15teams are thinking through like which
22:17systems can you put in place because
22:20this is only going to get bigger right
22:22and like we knew that hyperlane as a
22:24mission is way too big for one team to
22:27execute it's why we've taken this open
22:29approach it's why we've encouraged
22:31outside contributions that eventually
22:33led to like the mitosis team building
22:35the km wasm version the eclipse team
22:37starting off with the salana version the
22:39movement lab team the movement Labs team
22:42building both um uh Aptos move and S
22:46move versions of hyperlane um and again
22:50now other people building like their own
22:52modules but how do you extend that to
22:56not just providing the security but also
22:58like the additional eyeballs like who is
23:00going to be kind of like what I
23:01mentioned before with those uh this
23:03concept of like the Watch
23:06Dogs yep so I'd say we're still like
23:09still a lot of work to do um but we are
23:12kind of you know in constant elaboration
23:15with other people in the hyperlane
23:17ecosystem to like how do you scale this
23:18even further because these modules again
23:22like are to the benefit of everyone who
23:24uses hyperline like once one is out
23:27anyone can use it wherever it's
23:29operating and so we all want to have
23:31more of them and we all want to have
23:33more effective ones and so everyone who
23:36use like you're a chain that's using
23:38hyperlane like it's in your best
23:40interest for there to be like cheaper
23:42faster better like ZK modules and so
23:46shouldn't you spend if you have the
23:48capability to audit those or like help
23:51in the building of should you spend some
23:53time on it the answer is obviously yes
23:55uh but how can you sweeten that
23:59Arrangement thinking through I would
24:01maybe push back on like if I was like a
24:05rational hacker or a rational attacker
24:08which is like a term that someone used
24:10on a space just a little while ago um if
24:13I'm a rational attacker and I see that I
24:15can get I can get X from exploiting this
24:21and I would get y from an immuni bug
24:24Bounty and X is bigger than y then I
24:26would choose to attack the protocol
24:28rather than contribute to securing it so
24:31I think it first for sure like oh sorry
24:33I didn't mean to cat you up well just
24:35that from the incentives perspective
24:37this is like the cost of corruption like
24:40this would mean that if there is a
24:43higher cost to corrupt it than to
24:46protect it then we would
24:48expect an attacker to to do whichever
24:52one's in their economical best
24:54interest you to you would for sure now
24:59the best part and this the first time
25:01I'm like pretty happy not for the first
25:04time I've been happy about like the
25:06presence of a justice system most of my
25:08life that's a pretty you know Western
25:10Civilization is pretty sick I kind of I
25:13think it's awesome um I love Living in
25:16America I love having like a pretty well
25:18functioning despite all the people who
25:20have complain about our Justice systems
25:21like generally is like well functioning
25:25exceptions there's legal in that
25:27incentive calcul correct yeah so like
25:30the cost of corruption folks we will
25:32talk about that you know depending on
25:35like what let's say in like a proof of
25:38stakes there are things that you can do
25:39that actually look like corruption or
25:41clear like forms of maleant that don't
25:44fall under like they're not actual
25:47crimes per the Justice luckily you know
25:50it is stealing money and um the justice
25:55system has been pretty clear and I think
25:58that's why you've seen pretty much if
26:00you look at like the last 18 to 24
26:05hacks um they are all from people who
26:07are completely outside the system
26:10meaning like folks who that system does
26:11not recognize in any way right like the
26:13let's say Lazarus groups that's a
26:15criminal Enterprise that is completely
26:18outside the bounds of the system so of
26:20course like whoever they can attack they
26:22do but like the you know the standard
26:27hackers who like maybe in 2018 19 2020
26:30even would have like taken the Money and
26:33Run now they realize that like that's
26:36very very difficult to do and so that's
26:38why you're seeing all these negotiations
26:40you're seeing this like hey I have your
26:42protocol hostage I demand you know uh
26:46you mentioned right like I can get X or
26:48Y I demand y to be like some percentage
26:5310% um in that way converting them into
26:57like if you want to call them white hats
26:59because they are kind of still doing
27:01something kind of nasty but um yeah
27:04converting them closer to the path of
27:06the light with uh with
27:09dollars and and so you have to still be
27:14like uh those extra judicial actors
27:18right because like they're completely
27:19outside the system they will always go
27:22for X because there is no point for them
27:25with Y uh they just have a complete
27:28the thing with them they were never
27:30going to like support your security and
27:32so they're just like you know there are
27:34bad guys in the world and I'm not like
27:36here to you know decide on their morals
27:39but like they do things that we would
27:41consider to be bad things and you'd have
27:43a hard time changing them I think with
27:46the emergence of um the justice system
27:50into our space right like whether we
27:52want it or not that has basically helped
27:55redefine the bounds properly such that
27:58if you don't want to deal with that like
28:00now you're much more likely to think
28:02about y than you were about X almost
28:04irrespective of the Delta because like
28:07the calculation is now it's it's not
28:09just oh I can you know I can get y or I
28:13can get x x is much larger I'm G to get
28:15X so I can get X and then I have a cost
28:18associated with it which might look like
28:20you know a certain uh prison somewhere
28:23in United States like you're not a place
28:26you want to go to yeah or ending up in
28:29like a non-extradition country or
28:33but who knows who knows where you're GNA
28:36end up somewhere in like the buling or
28:38something yeah so it's basically what
28:40can you like what can you not change
28:41like we're not GNA be able to change the
28:44way that people like Lazarus or you know
28:46a focus of their capabilities and
28:49motivations are and then how can we work
28:52with the ones that you can change right
28:53someone who maybe like two years ago
28:55would have done something nefarious but
28:57today is like much more incentivized to
28:59just like notify you of the ISU give
29:01them a big Bounty because those people
29:03are actually really useful right like
29:05immuni has been useful for
29:07hyperlink into you know thousands of
29:10well yeah I mean I think you're you're
29:13either you're even furthering um this
29:15idea behind like like security for um
29:19interoperability and not so much just
29:22like with the with the legal
29:23ramifications but like the incentive
29:25model but even still like I'm going to
29:27press some even further like I think
29:29this idea of like unified API is not
29:33necessarily as strong as like unified
29:35security because security kind of forces
29:38liquidity fragmentation so you know
29:41let's say you know let's throw it back
29:42to like Harmony Harmony got hacked and
29:45even before they they were hacked the
29:47bridge was sketchy people were kind of
29:48sketched out about it um and and so you
29:51know the WRA tokens there were a little
29:53bit lower like usdt was 0.98 or or a
29:56little bit lower uh the yields were
29:59higher right because you have to
30:00incentivize people to come over um and
30:03so then you kind of had these like
30:05pockets of liquidity across different
30:07chains based almost on like the pricing
30:09of the wrap tokens as well as the yields
30:11based on like the the the security um so
30:15I'm not sure that like this like unified
30:17API of all chains is is is as good of a
30:20model as like a unified security
30:23model for sure like the I think you know
30:27you can make make like good arguments on
30:31the if you have something like unified
30:34security say like shared Security in one
30:36pool across everywhere it does mean that
30:41contamination right contaminates the
30:43whole pot and so if you think about like
30:46what the modular movement represents we
30:48started this off by talking about like
30:50what modular expansion really
30:52means if we were going to deal with say
30:55like a well-known wellestablished number
30:57of chains call it like couple dozen
31:00something like that then you could you
31:02know you could Envision some like
31:05standardization emerging and like we
31:08could all be civilized within it we
31:09could police that really well but the
31:11modular movement is creating basically
31:13like a complete open border policy
31:15anyone you have a chain and you have a
31:17chain everybody gets to have a chain and
31:19so in that one right like the concern
31:21that you first raised is like the
31:24correct one in that shared security like
31:26either I create like a very very
31:29difficult call it like immigration
31:30standard which is oh if you want to join
31:33this pool like you got to be this tall
31:35you got to have like this good grades
31:37you got to have like this track record
31:39Etc which is of course going to be like
31:41a very limiting factor uh for the
31:44modular ecosystems growth but at the
31:47same time like is kind of necessary
31:49because you don't want just anyone to be
31:51able to come in benefit from this pool
31:53and then also contaminate it and so the
31:58the biggest you know reason for us to
31:59opt for this model is well you want to
32:02give people more sovereignty then like
32:04they can't just be immediately opted
32:06into shamer pool and you have to Silo
32:08the risk such that again like you know
32:10One Bad Apple because there's G there's
32:13going to be more than one bad apple we
32:14already know that right like not even by
32:17intention just by virtue of like we
32:20talked about you again more security
32:22incidents have happened because of
32:24well-intentioned people making mistakes
32:26than there have been because of someone
32:30maliciously you know creating something
32:32to rug people like the ruggers most
32:34often are like hackers that are
32:36exploiting well-intentioned but
32:40um and so I think that's like that's why
32:45for this ecosystem in particular I think
32:48this model it's true it has its faults
32:50but we just have to build around them
32:52and kind of like make do with them
32:55because the alternative of just a
32:56completely shared model
32:58it's too risky and then everyone is
33:01[Â __Â ] even yeah like even the best
33:04actors would be [Â __Â ] in that model
33:06because they're all sharing the same
33:07like basically you know you don't want
33:09to Poison the Well yeah so you're
33:11permissionless but because you're
33:13permissionless the trade-off is you have
33:15to um Silo this risk and or basically do
33:18some sort of like permissioned
33:20deployment of a sense where you guys
33:23have a team of reviewers who review
33:25these things but then also in each
33:28individual component cannot affect the
33:31rest um yeah and basically like people
33:33have to opt into communicating with you
33:36right like you know in that sense I
33:39could get you know if I wanted to accept
33:41assets from say like if you wanted to
33:43accept assets from my chain and say like
33:45my chain becomes contaminated then like
33:48only once you've opted in can you like
33:50you know get get this sickness right
33:53here can get this contamination U but
33:55the fact that like I have contamination
33:58until you opt into that like you're in
34:00the clear um and so right like no free
34:05lunch as they like to say call that the
34:07E Denver cold don't want
34:12that the modular day after party cold H
34:15people didn't opt into that one you know
34:17that was that was a share we all shared
34:20in the same security of being in the
34:22same space and you know a couple of
34:24germs walk in and let before you know it
34:26everyone leaves uh with five days of
34:31decommission I had another analogy but I
34:33think that one was even
34:36better I wasn't you know Andy brought it
34:39up I wasn't even thinking about
34:41that well I think I'm all out of uh
34:44questions to grill you with John i' I've
34:46I've checked all my lists off Rob do you
34:48have anything love it my my last
34:50question was gonna be an analogy to uh
34:53cross margin lending markets and
34:56isolated lending markets if that's a
34:58fair analogy but well and and yeah that
35:01feel free to answer that question but I
35:02think the eth Denver shared security
35:05model is even even more accurate of an
35:08analogy so on the on the cross margin
35:11one like it it is similar in the sense
35:18um so if basically like you chose to if
35:22the asset you chose to accept as like
35:24your cross margin collateral if that
35:28shitty or unavailable for whatever
35:30reason then the entire system is [Â __Â ]
35:32now while it's working and while it's
35:35great everything is is you know
35:37everything is wonderful because we can
35:40Galvanize around this like uh singular
35:43you know margin asset and that's very
35:45convenient but the moment that that
35:48falls off and so like that's how you
35:49could have you know that risk not being
35:53siloed can affect the entire system and
35:55that actually like is why you know when
35:58different venues like which extend the
36:01margin to their clients like you know I
36:03used to work in Morgan Stanley like our
36:04Prime brokerage one of its key things
36:06was trying to have like a central risk
36:08book to understand like what are our
36:11risk you know like we are we are
36:13extending margin to all these clients
36:15and like we have to understand what does
36:17our risk look like at any point and when
36:19as soon as you start spotting
36:21contamination like they basically do the
36:25it's kind of like equivalent to
36:26amputation it's like that
36:28there's something wrong with that client
36:30cut them off um makes sense and so in
36:32crypto that's like it's very difficult
36:34to design a permission system and do
36:36that at the same time because like you
36:38lack the ability to cut things off in a
36:41coordinated sense uh and so it's
36:43actually like no it's a I would agree I
36:45would say it's a pretty good
36:46analogy got it yeah the idea of a
36:49contagion I think makes a lot of sense
36:53too all comes from disease models so you
36:55know that's why the Denver one made
36:58yep yeah yeah I think relatively healthy
37:01after Denver I don't know about you guys
37:02maybe I me too I gotta say the immune
37:05system you know you go to enough
37:06Denver's uh you know you're just build
37:08different by you learn you learn how to
37:10operate it's like when you're
37:12interacting cross chain on hyper You're
37:14Gonna Learn which modules to choose
37:15that's right yep that's right John I
37:18think I got one one more which is that
37:22um uh some of these some of these
37:25contagion can start to start to arrive
37:27when you're um when you're communicating
37:31these messages and if you do have a
37:33chain or security module that gets
37:35compromised um theoretically and I
37:38realize it's still so early ZK is
37:41supposed to solve that so that there is
37:43no uh contagion that can happen when
37:47you're sending something with a proof is
37:53hyperlane adopts ZK technology and goes
37:59like uh yes but basically like the way
38:02that that would happen is so hyper like
38:06already has uh so just last week one of
38:10the uh a module that kind of wraps the
38:13succinct Pro kind of reached its code
38:17phase um and previously there was one
38:19from electron Labs um so in that sense
38:22like ZK is available through the hyperin
38:24system where the limiting factor is well
38:27where is that module running and so
38:29again you know like yes it's running on
38:32ethereum so like you can use outbound
38:34for eum but is someone running it on
38:36like you know Andy's rollup not yet and
38:40so like once some ZK technology becomes
38:45completely ubiquitous and is so easy and
38:47cheap to run that it can be run
38:49everywhere you can definitely see that
38:52becoming like shared within the scope of
38:55that module and people could have even
38:56like silos of of that if they wanted to
38:59but it becomes far less like imperative
39:02to do that and the nice thing is like
39:04right that's not happening tomorrow but
39:06it could very well happen in the next
39:08five years and so with hyperline we
39:10wanted to design a system that people
39:12can build on without being concerned for
39:16like well in a year's time two years
39:17time three years time as the
39:19state-of-the-art advances do I have to
39:21have do I have to rip everything out and
39:23start from scratch or can I just keep
39:27everything I have use the same
39:28interfaces and then hey I was using this
39:32module over here and now I can use that
39:34module over there and basically like
39:36that that uh change process looking as
39:39simple as like updating a single
39:42contract to like change it to different
39:44address or you know redeploying uh like
39:48a new version but either way like
39:50keeping everything else the same uh
39:52because we definitely like you know we
39:55are big Believers in this
39:58technology uh I've spent a lot of time
40:01on the ZK front both like seeing the
40:03hardware that accelerates uh as well as
40:06like the latest on the software side it
40:08is coming like there it is sort of like
40:11inevitable it's but it's the same way
40:13that like you know 1995 I remember like
40:17we had a 64 megabyte hard drive trying
40:21to install like the first command in
40:23conquer was like what it's 40 megabytes
40:25this is bananas it's going to take up
40:27but 3 years later like the storage space
40:30expanded by an exponential relationship
40:33and now like you could go ahead and you
40:35could buy like a terabyte um you know
40:38like SSD drive for like basically
40:41nothing so that is all to say that I
40:43think by the end of the decade ZK is
40:45going to be much more ubiquitous and so
40:48we are trying to build a system so that
40:50it can adapt to that um in with like
40:54minimal disruption in practice does that
40:56mean there's a modular consolidation to
41:00a single security mod there might be
41:04it's like it's through time you know how
41:05like um a lot of people will say that
41:08like all of Tech has been like the
41:10bundling and unbundling of different
41:12services and so like right now we're in
41:14this great unbundling where like things
41:17are you know separating and now we need
41:19to build a connective tissue to do that
41:22but the next phase of growth might be a
41:23consolidation and then a further
41:25unbundling and a further consol
41:27validation um I think actually if you
41:29study you know the history of you do
41:32actually see like a lot of these
41:34bundling and unbundling or like
41:36expansion and consolidation
41:39Cycles that's it I'm
41:42out yeah this was great guys this is a
41:44lot of fun anytime I'll come on anytime
41:47you guys are the best