Noir 1.0

Transcript

Good morning GMGM. Thank you everybody here for taking choosing to spend your time at our third inaugural no >> bey as I was told that's how you GM here in Argentina. Uh lovely meeting you all. Uh and GMG GMs to all the English speakers like me. Um I am Savio the product manager of Noir at Aztec Labs and welcome so much to the Noon itself. Uh we have had some great workshops today that went through Oh, this is not intended.

Uh one second, let me make sure that's down. Yeah, resting your eyes are always good. Do remember that though. Yeah, we have had some great talks and workshops today that gone through parts of Noir um next story parts of Aztec and also a lot on the privacy community side of things. Um leading to the afternoon sessions, I'll be kicking it off with a brief talk, a brief overview of what's coming with Noah 1.0 itself. Um you might have heard of Noi uh for the first time actually. Uh who here uh this is your first Noon for you? Is this your first Noon? Can I get a hands raised?

Lovely. Wow, that's a lot. That's a lot. Second time. Oh, quite a few as well. Um, I will skip the third one uh onwards, but I Oh, okay. Uh, but yeah, despite how many times you have been to Noon, welcome to the Noah community always. Um, there are a lot of exciting things brewing both inhouse as well as as well as in the community itself. And so, yeah, let's just dive in. Um, but before going into the actual like 1.0 know itself. I do actually want to invite you all to a journey back to the year of 2020. Um, who here has been in crypto since 2020 or uh before 2020 even? Nice. Quite some

OG's around. Um, but yeah, if you can remember 2020 is the year where Eve 2 is still a topic of discussion. Um, which is which was superseded by proof of stakes or the merge if you have heard of it. Um but yeah back then people are still discussing uh the possibility of E2 itself. It was also the year where Polka Dot you might have heard of the chain itself actually launched their mainet um which yeah time flies in this space and 2020 was also the 2020 was also the year of DI somewhere if you were around. Um so curve for example launched that year unis swap v2 was launched that year. A lot of

exciting things were happening on chain on Ethereum and that also brought to an observation where well with all the interesting things happening on chain they are all also fully public on Ethereum itself and outside of Ethereum or maybe on tangent with Ethereum. Uh this was also the early days of CK. So this was actually a commit that I pulled from that year uh from circum itself. Uh that was how the docs look like back then. And it was roughly also the years when pl is plunk is g gaining momentum um as an academic work a bit of implementation here and there but yeah that was the year of 2020.

So this brings us to basically two questions in the year. First of all couldn't it all have been private? Why should all the interesting things happening on chain must be fully public, fully transparent for everyone? And couldn't the actual privacy works or the CK works be easier? Like yeah, circum is nice, but also it's like circuit crafting language. Plunk is great. So you don't have to get worry about like trusted setups, but there was barely any implementation if at all. Which brings to the birth of Noir. This was actually the first commit uh in the Noir repo in 2020. And with one very simple goal like we simply want to do

privacy. We want to make privacy simple to do. We want to make privacy simple to use. Um so privacy gets everywhere. Which brings us to everything everywhere. Throughout the years uh there were a lot of development that has gone into Noah itself. Uh from the first commits to the first version of Noah that we released. This is Noah Alva that we announced at uh Defcon Bogotaa. And we also announced uh Noah beta that came in the year that follows that was in def connect.

And of course we also released or pre-released Noah 1.0 last year at ifcon defcon Bangkok. Oh god you have come up with some better naming. Uh but yeah throughout the years there were a lot of changes that accumulated in the language itself. We started with the lovely fields that we all love uh that always overflows. Um so we also added integers uh where we can then bound numbers to actually work with also control flows what you might be used to in other languages if else for loops. Um but also a lot more fun even funner stuff like operations you can do comparisons. uh we have unconstrained functions which is

totally unique to the CK land. Uh most things are constrained but sometimes you also want to jump into that unconstrained world to get some uh additional performance. Guam from the crowd would actually be doing a workshop over it uh later today as well. We also have generics. We also have traits uh a lot of fun stuff from other languages like Rust or uh languages that you are familiar with that could be helpful in building whatever you want to do in CK. Uh we also have profiler. This is actually Maxim uh talking about and sharing about the profiler earlier this year on how you can actually visualize your performance of your no programs which helps a lot on actually diving deeper and optimizing your programs. And these are just like a brief uh a brief

look into all the things that have been accumulating on the noir side in terms of what you can actually do with it, what it can actually help you do as well. And it's not just everything, it's also everywhere uh that involves noir. So uh obviously Aztec has been using noir as the fundamental language to powers both its throw up but also its smart contract. But we have also been maintaining the solidity verifier for no programs. So you can also deploy your Noah apps on Ethereum or all the EVM chains alike.

And the arbitrum team has also been working on a stylus verifier for you to actually deploy no apps on arbitrum stylus. The stet team has been working on a cyro vera. Uh the stellar sorban team next door actually are also developing a sorban vera for you to deploy on there. Um the field team has been working with hashclo on a uh sway verifier. Um the Solana team is also building a graph 16 back end for you to use Noir on Salana. Uh the world team is building a proof kit for you to maybe even use it on the world chain uh at some point. But yeah, there are a lot of places where you can use Noir onchain.

But at the same time also offchain as well. Uh BBJS has the verify function where you can basically verify any Noir proofs in a browser or in any JS environment. Or if you fancy uh verifying on mobile, there's also the Mo Pro team. Oh no, the slide cut out. Uh but the Mo Pro team has also been working on a kit for you to actually do native mobile proofs but also for a vacation uh on your mobile phones. And there are also a lot of actually interesting example that we can see that works fully offchain as well. The CK passport team is doing a serial knowledge proof of your passport that can totally work offchain on a server client relationship or even P2P. Uh there's also the CK email team working on CK email primitives for you to prove

your emails which can also be verified offchain. Uh there's also the Fedlabs uh team working on Terry escape which is a fully offline P2P game uh that is played offchain as well. There's also the open bands team doing a social platform where you can submit your knowledge proof of yourself onto the platform to testify kind of your pseudo identity and make social posts around it. So with all the different places and all the different features, Noah sets out to basically do one simple thing simply privacy but now also privacy everywhere which leads us to the time that we are living in now. Um throughout the last few years things has changed a lot uh

both within the privacy space but also uh in the entire digital space as well. uh we have been seeing for example the Estonia government Estonian government putting more e identities on chain uh as an example on the identity side uh the Bhutan government has been working with EF on bringing their identities national identities onto Ethereum as well through the initiative uh EU has al also been pushing the EU digital identity regulation framework to bring more identities on chain so a lot of identities are getting on chain at the same time We're also seeing a lot of well we call it real world assets but uh for real world people that's probably

just assets bringing real world assets on chain. Uh Western Union has been launching their stable coin. Japan has been launching their stable coin. Uh even Argentina or Latin America has been launching their stable coins. Uh Canada has been bringing stable coins. Uh the US has been bringing stable coins on chain as well. So we are seeing a gradual shift from momentum of digital digitalizing a lot of uh what we might have been seen as important stuff both on the identity side and also on the financial side and with all the failures coming on chain and coming to the digital world. Uh in parallel to it we have also been seeing a huge growth in the privacy sense of things as well. Of

course in our web 3 community we have been seeing a lot of traction going on uh catching up on on the privacy side. We have seen like rail guns uh deposits continuously flowing in. We've seen ccash shielded deposits flowing in as well. A lot of interesting development on well both on the noir side but also across a lot of different other privacy projects taking space as well. And uh with even outside of the web free space that we are in uh we're also seeing a great traction of privacy awareness as well uh with the UK's introduction of the age for vocation uh regulation we have seen like for example ProtonVPN picking up like 14 times uh

traction growth just throughout those few days. Uh we have also been seeing uh well the EU regulators planning to impose chat control or uh basically encryption with back doors uh in your chats or messages and we've seen we have been seeing a lot of like grassroot momentum in pushing against that and this is also a uh study or survey done by Deoid the uh consultation firm that you might have seen. Um, it might be hard to read from afar, but it was basically asking like how willing or how reluctant are the general folks being willing to share the data with companies or services. The green part is like I'm willing and the gray part is basically

very unwilling or well somewhat reluctant to share to share data with those services. And we can basically see like it grows from different facets but the majority of the users aren't actually uh happy about sharing more data with the applications or the services that they are using as well and hell even our favorite YouTuber PewDiePie has been shilling about like Linux how do you actually get to graphine OS who here knows who PewDiePie is actually okay uh for the ones that did not have your hands raised congratulations uh on your lovely life that you Uh to the ones that actually raise your hands, congratulations to a lovely life as well. But after this week, do go touch some grass, please. Uh but yeah,

we have been seeing a shift of momentum on how people are actually caring about privacy as well. Like it's not just us, you and me shilling about privacy throughout the years, but people not caring about privacy is just a talk from the from the previous century at this point. Which brings us to the next chapter. What have we been cooking in the kitchen itself? Um you might have thought you might have thought of or had impressions of like what the no team does. Uh but what the no team does well most of us things uh we just like crash our laptops and uh get mad at it. But what we actually do have we have been cooking a

lot of things inhouse as well um to basically close the last mile when it comes to uh privacy itself as well as bringing privacy to people's hands. We have been for example fussing no other language uh if you are familiar with fuzzing it's basically well you can in one liner is like randomized testes randomized generated testes for your codebase. So we have been fussing nor the language. We are not fully there yet. As you can see we have a uh tiny red cross there that we still haven't cleared out but uh we have been making progress.

We have also been creating a fuzzer for people to actually fuss their own no programs as well. So for example if you are building a no project if you are planning to deploy it into production strongly recommend you to actually fuss test your program as well to make sure there aren't caveats uh that you might not be aware of. We have also been working with the ray labs team on getting uh well starting off with no programs being formally verified uh through the lamp project. Formal verication if you think of it um in simple words it's like mathematically proving that your program is right uh in addition to well the traditional side of like fuzzing testing auditing etc.

And we have also been tightening up our way of um actually gathering vulnerability reports from the community as well. We have recently set the GitHub security advisory uh process up on our GitHub. So anyone can actually report privately to us on uh bugs or vulnerabilities that they have come across which is helpful in actually responsibly uh patching the bug as well as privately communicating it with uh uh projects that might be affected first and hopefully minimizing whatever uh damages it could bring. And we have also been working very hard for the last few months over internal audits of the code phase as well. uh we have splitted the code base up into I think at least 10

groups and this is like one group only. There are a lot of work that we have been going through on making sure we are doing what we expect it to do uh when it comes to no itself and we also acknowledge that perfect is always the enemy of good like we aren't striking for the perfect security we aren't striking for the perfect version of noir before calling it 1.0 you know, but rather we are taking the approach of gradually and progressively securing Noir uh in-house as well as um with the community as well. So with the Noir 1.0 launch coming up, we do also anticipate two more external audits uh which we have already kickstarted but has been uh

progressing through on no the language to bring further uh kind of stability but also confidence in the codebase itself. And we also anticipate a lot more security initiatives to come. Um for example, competitive audits or open audits that you might be familiar with or buck bounties as well. Uh that helps uh facilitate white hat uh reports as well as um auditing of the codebase. And we are here somewhat we are working towards uh the 1.0's launch and uh quite a few of u yeah security initiatives that uh we'll be pulling through.

which basically paints out the timeline that we are working against but also at the same time um if you are actually building in noir or if you are interested in building in noir then what are the ways that you can actually participate in this timeline not just helping us but also simply using noir for uh to achieve whatever privacy uh initiative or privacy application that you would like to use or do uh it has been quite a journey since well no ala no no beta is it coming up no beta uh noir pre-release and the current version of no has been quite a journeys throughout the last five years developing the language but how we see

1.0 know it's it's the official starting line for we to actually raise towards privacy to actually raise towards uh uh production readiness uh privacy applications as well and uh there are three immediate kind of things that you can consider working through. Uh well of course you can you're always welcome to stick around for Noon. We had some great talks earlier but we have even greater talks coming up uh later today. Gian will be talking about Noir performance. how do you optimize your no programs uh at a lower level uh in a workshop here and then uh the more pro team would also be talking about how do you actually do native mobile proving with no later today in this room as well uh next door

in the Aztec room I believe what's happening right now is uh the diva wonderland is going through for example how do you code your first Aztec smart contract with noir um in the community room what's happening right now is the stellar team is also talking about how do you actually how can you actually deploy deploy your no apps onto Stella Soraban as well. So a lot of amazing workshops and contents uh coming up in this afternoon still in this venue but also at the same time uh this is probably the best group of builders that we have gathered in uh in this venue as well. So there are a lot of people for you to like brainstorm ideas against to hack around ideas against to build ideas around against as well. Uh so do tap people's shoulder and uh start chatting as well. And the upcoming EF global

throughout the weekend there's also an Aztec track where you can hack around as smart contracts or uh things that you might have been dreaming of building on Aztec. Uh that would be a good opportunity. If this week is not the week, there's also the cyber punk hackathon that is happening online uh from mid this month all the way to late this month or well December one. Uh there's also an Aztec track up there where uh there are also a few ideas on what you can actually hack around. So if this week is not the week, next week could also be your week as well.

And with all the things that we have gathered for 1.0 itself, uh as well as what the community has been building plus the like new era that we are living in now where there's more privacy awareness than ever, there's more demand for privacy than ever. Now is the prime time to actually build our privacy apps uh for our parents, for our grandmas, for our friends, and to bring privacy to this world. Um, and it has been an honor for me to actually have worked on the project for so long, not just with the team, but also with all of you on bringing privacy to this world. Um, and I do look forward to seeing you around this place as well to uh brainstorm

ideas together, to hack around with noir, to hear your thoughts on Noir as well. Um, and it has been my honor to present you Noah 1.0. And before we go, uh, Noah 1.0 will be dropping 2026 Q2 next year. I'm Safio. Thank you all. We have time for questions. Like if a person wants to contribute to nor like from outside apps a tech of course do you have like a to-do list or

or a place in which you have like a yeah we have these problems we need this tooling please help us to build it uh where can I find >> of course um this list has always been an evolving list at some point but yeah there are definitely a few things that you could do depending depending on like your interest uh or well your proficiency in no or in the stack itself. Uh for example, simply improving no's documentation. To me, I would say that's already like a huge leverage that we can pull. Um there are a lot of content covered in the documentation written by me myself, but also a lot of the engineers on the team. Uh which might not always be the easiest way to

consume the the knowledge itself. So if you notice anything we are always well noir is open source from the ground up. So we are always welcoming PRs on improving for example documentations or reads for example and there are also uh if you are more like cryptography oriented or uh more interested in the for example mathematical or algor algorithmical side of things um then libraries might be a good one as well. There are a great set of libraries today already with uh for people to build out what they want to build, but I would imagine there being 10 times more ideas that could only be powered with certain libraries in place. Um so if you have certain ideas about libraries that you

would like to build, feel free to build them and share it in the Noah discord, share it on Twitter. We are always happy to amplify as well as um share thoughts, cross thoughts with you as well. And if you are more a uh I guess software developer oriented personnel maybe less crypto interested or even but would also like to make some impact in addition to just stocks I would imagine tooling being a very good part uh and very uh uh underexplored part at this point as well um for example if you make things easier for people to import and export libraries if you're making things for people to more easily discover for different things happening in the Noir

space. If you are making certain like manual interactions with Noah nowadays a lot more automatic, those could all be like great ideas for like tooling that could exist and help the bigger Noah ecosystem as well. Um but fundamentally I would just recommend you to like try Noah out, talk with people in the space and see where the gaps are. Uh those would be the the fundamental principle for what you can actually help the bigger ecosystem to grow together. Well, how about small field support? Is it on the road map?

>> It's not officially on the road map yet. Uh but that's definitely something we have been considering for next year. Uh we have been very focused on shipping 1.0 first. Um but yeah, small field support and a bunch of other features we have been discussing internally has also been uh growing up on the list itself where we would be able to tackle hopefully next year after shipping 1.0. I participated in the previous noir hack and it was really cool. Sorry to put put you on the spot, but is there another noir hack coming?

I don't know yet. Maybe uh but yeah, we'll share more if we do have plans. Thank you. >> Okay. Uh who do you think Noir is most suitable for? Which kind of profile and something like that? >> Right. Uh I would say if you are a developer or at least interested to become a developer then and if you are interested in doing privacy apps Noir is for you. Um it's super easy actually to to code noir like uh I've briefly talked about circum um I briefly well I've

briefly mentioned other CK stack as well but a lot of people like come to me asking about like what do I need to know to start coding in no um I've always been telling them you only need to know like very fundamentally how cate proofs work like there's a prover there's a verifier and basically it's an interaction between them two once you get that it's super easy to start coding uh noir is like very rustlike. If you're familiar with the language or even not familiar with the language, it's quite easy to pick up uh as long as you are already somewhat comfortable with programming itself.

>> What will be the tradeoff with more classical like ROS 16 or CGS libraries that allow you to do ZK proof but without like the language layer is there tradeoff, >> right? So the way that Noah actually is architectured is uh there's no the high level language on top but it actually works with different proving backends. Well the caveat being some small field back ends might be trickier to work with but the architecture is aimed to be worked with different proving backends. So uh if you have seen for example Burtonberg the proving back end that's the proving back end that uh uh the Aztec labs has been maintaining it's based on ultra honk the uh well the proving technology but for example what

the Solana team is building it's actually a graph 16 proving back end for you to use as well so you can actually switch between like different proofing backends for different well tradeoffs and to your question on what are the trade-offs 16 is probably the golden standard when it comes to like proving speed uh proof sizes. Um but the biggest draw down that it comes with is that you have to do a trusted setup or a trusted setup ceremony for each basically version of circuit that you are you are building. And uh the problem with that when it comes to actually deploying applications or projects in production is well doing a trust setup every upgrade might be somewhat reasonable. So like if I'm upgrading from v1 of my app

to v2 I do a ceremony for with my users and bring things forward but at the same time if you're deploying into production uh patching your app is also like super important and I don't think any team would actually like to for example whenever they come to a small bug in the circuit and they want to patch it they run the entire setup again with all the community involved. Um so that's why the plong technology came to place and gradually evolved to what ultra honk is right now. Um for some specific use case it might be slower in terms of proving speed. The proof sizes is also a bit bigger but the beauty is that you no longer have to worry about a per circuit trust setup. You just need one trust

setup that we can use from long ago uh where a lot of people are already involved to secure it and call it a day and start building your app deploying your app. So yeah, that's roughly the trade-off between those two. There are a lot more other proving technologies as well. If you are curious to learn more, I'm happy to talk more off stage. >> Hi, thanks for the presentation. um what kind of tooling you think could be helpful for the Noir ecosystem and do you think it is necessary to have some tooling related to Noir with a specific blockchain >> right that's a good question um I can

start with some I guess brief ideas on on simply vanilla noir tooling um we had an early version of uh the debugger but it has also been falling out of date so getting the debugger up to speed uh to the latest version would al would definitely be great on actually like going through your programs debugging your no programs. Um I've also gradually been hearing more about how difficult it is to discover and make use of different libraries as well. So I briefly mentioned it just now but for example a very nicel lookinging package registry with a package management uh system could be nice for people to actually know like what are the things available how can I actually use them or more

easily use them as well. And then uh there are also tooling on oh I lost my track of thought there but yeah those are some some of the like vanilla noir tooling that could be useful for the like on chain part uh to be fair it's already quite simple like the proving back end for example using burenberg as the example you have the burdenberg back end you generate a proof out of it out of it uh if you want to use it with a chain specifically you only need like a very simple either SDK or a CLI tool uh that well generates a verifier for the Burton proving back end uh deploys it and then helps people to

like I don't know transcode that proof into the smart contract syntax that is needed and get it verified on on that chain specifically uh that part like if you are a somewhat proficient uh cryptographic engineer would only take what a few days of work to get a verifier in place of course it would takes a bit more time uh to to uh beautify the thing to uh flesh out the thing but yeah uh it should be relatively simple to do if you are going across chains as well. Um so yeah if you have ideas happy to talk more off stage as well I think that's all the time we have but yeah thank you all for uh being here. Uh next up Giam will be doing a very nice

workshop that uh Jake would probably introduce so I will leave the stage to them.