Crypto AMA with AVA (8.30.19)

Guests:

  • Emin Gün Sirer

  • Kevin Sekniqi

Moderator:


Moderator:

Let’s give a warm welcome to Kevin and Gün of Ava Labs!

As a reminder for everyone participating—please keep the discussion respectful at all times.

@kevin @emin - could you start off by giving us a brief bio on your background as well as how you got started in crypto? And then a short overview of Ava and a brief update on your progress to date? We’ll then be off to the races with questions.

Kevin Sekniqi:

Thank you, Spencer, great to be here. Before I start, Gün is on the road but will jump in soon. So I got into crypto after discovering it on Reddit many years ago (soon after the whitepaper had been announced). At some point I decided to study cryptography in college, and cryptocurrencies during my PhD. 

As far as a brief overview of AVA goes, the shortest description I can put together is by saying that AVA is "an internet of assets". It's a decentralized platform, that enjoys high scalability and performance, yet provides a really flexible network model that no other platform does, which makes it highly suitable for creation and trading of new markets and assets. 

We've come very far in a very short period of time. We closed our first round of funding in February, delivered private testnet in May, and are looking to deliver public testnet very soon.

Moderator:

Thanks Kevin, this is helpful background. Let's jump into questions from the audience!

Participant:

How many people are part of Ava Labs, and is there a team page?

Kevin Sekniqi:

No team page right now, but we are 25 people as of a very recent headcount, although we are growing very quickly.

Participant:

Who are the researchers in Ava Labs?

Participant:

What is Emins involvement with BloXroute and are the two projects related at all ?

Participant:

Hello Kevin and Gün! Thank you for the time. Could you provide an overview of your staking economics?

Kevin Sekniqi:

We have five on the team currently. Prof. Sirer, Prof. Bernard Wong (UWaterloo), Ted Yin, and myself, are core engineering researchers. Amani Moin (Cornell, Johnson School) is core economics researcher. We have one more joining in soon, Phil Daian. We also have researchers that are not part of the team right now, but have been in the past and act as advisors: Prof. Robbert van Renesse (Cornell) and Prof. Ittay Eyal (Technion).

Participant:

Hey Kevin! Nice to see ya - hope Williamsburg is treating you well. Would love to hear your take on Perlin's criticism of Avalanche.

Kevin Sekniqi:

I cannot comment on this on behalf of Gun, but I will say that the involvement is minimal, if at all existent.

Emin Gün Sirer:

Hello everyone

Participant:

Hey Gün!

Moderator:

Welcome @emin and thanks for joining! Feel free to provide us with a quick background and we've already jumped into questions.

Emin Gün Sirer:

bloXroute was founded based on the code I developed at Cornell University with Soumya Basu. But I am no longer involved with bloXroute in any capacity anymore.

Participant:

Hello! I'd love to ask, how exactly on a technical level does Avalanche plan to support smart contracts in spite of supporting only a partial ordering of transactions?

Kevin Sekniqi:

Not fully sure that there is any criticism that is relevant! From my last reading of some text that they had written, they were saying something along the lines of "Avalanche can't do total ordering", but it was *by construction*. It was meant to do that to increase throughput. However, total ordering is trivial to achieve. We have an internal implementation of linear chains using Avalanche, which will be released on public testnet.

Participant:

Also, how is your resolution with snowball only supporting a small amount of adversaries?

Kevin Sekniqi:

Williamsburg has been fantastic 😄!

Participant:

Also, when would you plan to open-source the numerous language implementations you have available?

Participant:

Very curious why you decided to only provide your testnet to a private audience for the time being.

Emin Gün Sirer:

Hi Spencer, 

Thank you for having me onboard. My background is simple: I've been in the space for a long time, going back to 2002, 6 years before Satoshi, when we built Karma, the first cryptocurrency to incorporate minting via Proof of Work. 

With Ittay Eyal, we discovered the biggest known flaw in Nakamoto consensus known as selfish mining. I worked on securing cryptocurrencies through covenants, and we invented the notion of "vaults", which have been incorporated into a few systems by now. We invented Bitcoin-NG, we also did the fastest known Layer-2 solution known as Teechan/Teechain. This link provides some background into me and my research philosophy: http://hackingdistributed.com/egs/

Emin Gün Sirer:

We have extended the Slush/Snowflake/Snowball/Avalanche family of consensus protocols with a new variant called Snowman that creates a totally ordered chain.

Participant:

Do the consensus protocols developed by Ava Labs depend on a randomness beacon (such as RANDAO, Dfinity-style randomness, VDFs, ...)?

Participant:

Why was it then did avalanche choose to originally implement snowball then when snowman provided better guarantees?

Kevin Sekniqi:

So let's decouple a few things here and really break down some misconceptions (there were a bunch written on Twitter which I thought were so incredibly off-base). 

Snowball (for safety) supports an arbitrarily large and parametrizable set of adversaries. You can set this very high, by setting alpha --> k. For liveness, Snowball is the only voting-based protocol that can still make progress *without* a leader. However, a leader is an easy upgrade, and can be retrofit to build a linear chain, thus providing the scalability guarantees of the Snow-family, with the liveness of classical.

Kevin Sekniqi:

Because you might try to take our code 🙂

Kevin Sekniqi:

Avalanche and Snowman are apples and oranges. One is optimal for partial ordering. Another provides total ordering. Different applications.

Participant:

The code either way is described in the white paper - did you intentionally censor portions of the code then to prevent competitors?

Participant:

Snowball sampling is a very trivial process outside of consensus :)

Participant:

Why only tackle decentralized finance then and why not open source a smart contract SDK?

Moderator:

Kevin and Gün, I'd also like to talk about your platform architecture. Judging by your 2-pager, it appears that flexibilty is a key feature -- can you talk more about what that means and your plans for that? Will Ava dapps be able to leverage different consensus mechanisms?

Participant:

Given the diversity of consensus protocols (Slush/Snowflake/Snowball/Avalanche/Snowman), has Ava Labs decided to focus on a specific one?

Participant:

💀

Participant:

Is throughput the only advantage that Ava gives for finance use cases?

Moderator:

Also, quick gentle rules reminder for the audience: (1) stay on topic (2) be respectful -- thanks everyone, good stuff so far.

Kevin Sekniqi:

Great question! So Slush, Snowflake, and Snowball are binary consensus protocols of various security guarantees. They, by themselves, must be used as primitives. Avalanche is a DAG implementation which supports high throughput for commutative operations (eg. UTXO payments). Snowman is a linear implementation for applications that require serializability. The only final protocols to be used eventually are Avalanche and Snowman.

Participant:

Hi Kevin and Prof gun sirer. How does Ava labs make money ? / revenue model

Emin Gün Sirer:

Ava has a unique network model, totally different from the 1800 other currencies on coinmarketcap, which all seem to have copied the same model from Satoshi. Namely, they have a universe with a single coin, a single scripting language, and a single network. This makes it difficult for them to:

- evolve the coin, because they need to convince the people in the network to adopt their changes, which involves bickering with a large group of internet volunteers

- offer guaranteed services, because the foundational service that they offer is based on best-effort service, provided by volunteers. 

Ava is different. We envision having thousands of coins on top, in addition to the Ava token. We plan to support many different scripting languages. And most importantly, we envision having lots of subnetworks coexisting within the same overall structure.

So, for instance, you can issue Cole-Coin on Ava, and have it support bitcoin script + ZKPs + ring signatures, as well as the ability (say) for adjacent coins to merge. This might make sense for a real estate application. And additionally, you get to determine the set of nodes that hold the state for Cole-Coin, e.g. that they possess a certificate. We refer to these certificates by color, e.g. purple nodes support Cole-Coin. You get to determine who these purple nodes are, and what they have to do to get the purple certificate. For a real estate application, they might have to promise that they will hold the data for archival, for the next decade. 

This revolutionary network model enables differentiated staking services to be offered on top of Ava. Businesses can build on their own coins because they aren't relying on internet volunteer and lowest-common-denominator service. And the stakers themselves get to collect fees commensurate with the service they are providing. Stakers supporting a real estate coin should make a few thousand dollars per transaction, which is roughly what it costs to make a change in a permanent land registry in the real world. They can do this on Ava because of our unique network model.

Participant:

New chains have zero assets — Ethereum has ETH and a myriad of tokens. This makes DeFi applications like Uniswap or Maker interesting on Ethereum. How do you plan to compete with this given the headstart?

Emin Gün Sirer:

We are a tech company. After the initial bootstrap phase, our revenues will come from partnerships with others who want to issue assets on our platform, as well as services we plan to build and operate ourselves.

Emin Gün Sirer:

We plan to provide bridges to existing cryptocurrencies.

Moderator:

Really cool.

Participant:

In a recent (private) conversation Vitalik wrote that Avalanche is "vaguely Casper CBC-ish". Would you agree?

Kevin Sekniqi:

Nope! This is by far, the smallest value add. We've thought super hard about real compelling use cases for us, and our current model stems from months of discussions and market search.

There is a real huge problem in current financial markets: reach and liquidity. For example, Norwegian oil fund has $1.3T in cash, and even if they were to deploy it across all investable assets in the US, they would single-handedly bump up the entire global markets. A lot of really smart people are being paid a lot of money on Wall Street to find new ways to deploy capital. 

What we have built with AVA is exactly a product that solves this. It is a network-of-networks type architecture, where each network is feature-specific. It supports a specific type of VM, validators, etc, as well as its own set of compliance and trading rules. 

So you can have something super decentralized like AVA-ETH (essentially all of Ethereum running as a subnetwork in AVA), or a centralized subnetwork managed by a few trading firms.

Participant:

How is this different from what SKALE is building on Ethereum? Make-your-own-chain with a rotating pool of validators that meet certain spec requirements

Participant:

Awesome. Thank you, Gün!

Emin Gün Sirer:

Absolutely not the only advantage! Several differentiate us:

- Low latencies: on the order of a few seconds

- HIgh throughput: we hope to match or exceed Visa, something that our current testnet does routinely

- High Scale: Ava enables thousands to millions of people to participate meaningfully in the protocol, something that no other protocol can even come near. Even Bitcoin is just 19 mining pools, and classical algorithms croak at about 100 participants. Ava can incorporate many more.

- Revolutionary new approach to staking: I explained this above, we have a network model that gives rise to differentiated staking services

- Governance: Key parameters in Ava, unlike most coins, are subject to change dynamically. There are limits in place to ensure that they do not change too much per time unit, so the system is predictable.

Kevin Sekniqi:

You can launch AVA-ETH and wholesale import all these smart contracts directly. But while I'm a big fan of Ethereum, the current set of DeFi applications are incredibly tiny in value. We will be focusing on onboarding many new assets and markets.

Emin Gün Sirer:

Snowman and Avalanche go hand in hand. As you all know, "partial order" is sufficient to build fully-ordered chains, and one can embed a chain within a graph. So, we use Avalanche for building a giant DAG, with smart contracts forming totally-ordered chains within the DAG.

Participant:

Tiny relative to total addressable market or tiny relative to the blockchain space because ETH has pretty much the entire DeFi market.

Participant:

Does this imply snowman is another protocol that works separate from avalanche?

Emin Gün Sirer:

Tiny relative to the total addressable market. We want to go after assets that are not in crypto at all right now.

Emin Gün Sirer:

No.

Kevin Sekniqi:

A *lot*, but here's one to just kill off the comparison really quickly: cross chain swaps. Trading assets in different subnetworks is a matter of a second in AVA, but not on Ethereum/SKALE.

Participant:

Why wouldn't ETH capture this?

Participant:

Gotcha, so Avalanche as a whole has been hidden this whole time and Avalanche is actually Snowman?

Kevin Sekniqi:

It cannot. Both performance and architecturally.

Emin Gün Sirer:

Ava's performance, security and network model are, we believe, a perfect fit, while existing crypto has difficulty doing anything outside its own universe.

Emin Gün Sirer:

No.

Participant:

Sure it can, not at base layer of course, but no broadcast network will scale like that. ETH2 + L2 should serve this purpose equally well and already has the interest around DeFi. Would depend on a complete ETH2 failure imo

Moderator:

Can you talk about the security model of this hypothetical Cole-Coin network? Does it derive any security from a main chain or is it independent?

Participant:

Yeah I think the cross-chain stuff will be an interesting space to compete in with Cosmos' IBC, Polkadot parachains, and eth shards + L2's

Participant:

Care to describe Snowman more precisely then? As you answered both of my questions being a double negative.

Kevin Sekniqi:

To some extent (maybe in the voting procedure), but not quite. What the folks at ETH 2.0 are doing differs from the Avalanche protocol, at least from my talks with them. ETH2.0 is leaning on a beacon chain to orchestrate shards. It's providing a source of randomness which all other sharded chains use. Avalanche doesn't have this requirement. I'd say that ETH 2.0 has more in common with Dfinity's threshold relay than Avalanche. In contrast, AVA uses Avalanche at the base layer (which provides payments + swaps), and chains built on top are organic and feature-specific subnetworks/execution environments, not shards.

Participant:

@emin what will allow Ava to scale so much better when Bitcoin and Eth  both have scaling challenges that seem somewhat orthogonal to the consensus algorithm chosen?

Kevin Sekniqi:

Can you elaborate on those challenges?

Participant:

Why do you describe Avalanche to be an asynchronous protocol? All the way from the FLP impossibility theorem we know that we can't establish consensus under a fully asynchronous network.

Emin Gün Sirer:

Staking happens first in Ava, so that provides a base level guarantee for all nodes in the system. In addition, the assets issued on the Ava network depend on the nodes supporting that coin. That set is determined by the asset creator and can change dynamically. There is full transparency for all participants.

Kevin Sekniqi:

If I had a quarter for everytime someone tried to explain FLP to me, oh boy!

Kevin Sekniqi:

I'd have *many* quarters.

Participant:

well, the obvious one is every node is tasked with keeping / validating the entire state

Participant:

When are the papers for staking and economics to be released? A lot of people have been waiting for them to be out.

Participant:

That is not a correct statement of the FLP impossibility result.

Participant:

I guess I just can't see why, without absolute failure to launch, ETH2 wouldn't be able to capture these cases because of its head start. Even *if* the technology of the other chains is superior (which is largely unproven and untested).

ETH2 also has the ability to incorporate these ideas where proven.

I think AVA and the other networks attempting this would truly need to outperform ETH2 by 10-100x, and most importantly ETH2 substantially underperform in order to gain traction.

Kevin Sekniqi:

Right, so this is a pretty bad design on ETH, which is why it won't work for this new financial fabric we're building. Everything goes through one main pipe. Imagine building the internet where all requests go through one global router.

Participant:

As we know, Avalanche does not by default guarantee sybil resilience.

Participant:

Humans settle on perfectly imperfect systems every day as a natural consequence of it's what everyone else is using.

Participant:

Care to explain why?

Participant:

Asynchronous networks are a very rigid and obvious definition based on FLP impossibility.

Participant:

Especially based on the proofs.

Emin Gün Sirer:

Because we have a lot of projects that are just copying what we do (hello Kenta :-), we will make these public only when we make our code public.

Emin Gün Sirer:

Ava does.

Participant:

Why the pushback when all other crypto projects are unable to copy then?

Participant:

FLP impossibility states that you can't have guaranteed liveness and guaranteed safety. Avalanche doens't claim to have either.

Participant:

Yep, though Snowman claims otherwise.

Participant:

I want to be sure it's not vaporware.

Participant:

What's the timeline for AVA code availability?

Participant:

I'm sure Snowman also have probabilistic safety and probabilistic liveness

Kevin Sekniqi:

Very true! That is why QUERTY is the winner keyboard despite being obviously subpar to DWORK keyboards. But, Ethereum is simply not the right design. It: 

1. No built-in protocol-level compliant mechanisms. 

2. Bad architectural model for issuance of new assets (privacy etc). 

3. Bad performance for global trading. 

Etc.

Participant:

Will it be MIT licensed?

Emin Gün Sirer:

Absolutely. We do believe that what we offer is technologically lightyears ahead of what Wall Street offers today, and look forward to seeing new assets built and issues on our platform.

Participant:

@emin @kevin Snowman has probabilistic safety and liveness, right?

Moderator:

Please follow the rules: https://t.me/c/1477285922/3879

Participant:

I agree for ETH1.x but how does this apply to ETH2?

Kevin Sekniqi:

It's all probabilistic yep. Also, FLP is for *deterministic* and *Dwork-style asynchronous* protocols.

Participant:

Will do, pardon for that - want to validate AVA's technology.

Participant:

Chernoff bounds are the only thing that proves AVA's validity. Why not combinatorial proofs?

Emin Gün Sirer:

People often forget that the CPU inside their laptops is also a probabilistic device. It miscomputes with a very low probability.

Participant:

Snowball is a combinatorial sampling mechanism after all.

Emin Gün Sirer:

The world at large is probabilistic as well. For instance, there is a probability that all the oxygen atoms in this room are going to bounce to the far end, leaving me to suffocate in a sea of nitrogen on this side. This is a low enough probability event that I'm not worried about it.

Participant:

"ETH 2.0 is providing a source of randomness [...] Avalanche doesn't have this requirement" => How does Avalanche sample block proposers without randomness?

Participant:

Piggybacking off that, we've seen that the best tech doesn't always win. What is your guys' plan to woo developers to build on Ava?

Emin Gün Sirer:

All randomness in Ava is local. The protocol is leaderless and does not rely on a random number oracle or any other source of global randomness for its correctness.

Participant:

Consensus however is very well defined.

Participant:

does AVA's consensus assume specific nitrogen balances?

Kevin Sekniqi:

Great question! Best tech most definitely does not always win. It's all about the team executing. Part of our move to NYC was to get close to the guys doing trillions of dollars of trades a day, and really understanding their problems.

Participant:

So AVA is working on international trade rather than generic finance?

Kevin Sekniqi:

For the crypto community, if we did our job well, the dev experience should be *identical* to that of Ethereum, for those that choose to deploy on the EVM chains.

Participant:

People have had trouble with deploying contracts over EVM though, is there no effort to optimize the process?

Participant:

Blockchains are very esoteric to traditional devs.

Kevin Sekniqi:

Uniformly distributed nitrogen 😄 !

Participant:

What is nitrogen?

Emin Gün Sirer:

:-) Good question. Avalanche can actually handle unprecedented numbers of Byzantine actors! In fact, we can tolerate, with safety but without liveness, Byzantine nodes that exceed the 51% boundary.

Emin Gün Sirer:

The analogy got stretched a bit. I was initially using a physical example to point out that the real world is probabilistic. Now, we are referring to byzantine component.

Participant:

Is there any proof on this? What about bad apples abstaining liveness of the network by preventing a successful sampling iteration in Snowball as described by @Alex ?

Participant:

It sounds like Ava will have a VM. what tools do you plan to provide developers to ensure they can be confident in their use of Ava?

Moderator:

emin @kevin can you talk about future plans for scaling the team? How are you planning to onboard more dev assets, new hires you're looking for, etc.?

Participant:

Which hash function does Ava intend to use (e.g. SHA256)?

Emin Gün Sirer:

Can you describe in a bit more detail what kinds of tools you have in mind?

Kevin Sekniqi:

Maybe some relevant background reading on how you can trade liveness and safety https://medium.com/@kevinsekniqi/on-safety-and-liveness-trade-offs-in-consensus-protocols-23b9bbb61e38

Emin Gün Sirer:

We use RIPEMD160 and SHA256 at the moment. We mirror Bitcoin in our use of hash functions.

Emin Gün Sirer:

Except, of course, there is no mining.

Participant:

formal verification tools? a set of libraries that do common computations?

Participant:

Which signature scheme do you use?

Kevin Sekniqi:

Base AVA has a super simple VM that resembles bitcoin script. Other chains have their own VMs, imported as plugins.

Emin Gün Sirer:

ECDSA

Participant:

Gotcha, so is liveness compromised then? I just really want answers to be specific here to answer my question. If there is both, I would love to know more.

Participant:

👍

Participant:

If you have a section in your whitepaper, please refer me to it. I'll have a glance through.

Kevin Sekniqi:

I mean happy to give you lectures on consensus and stuff, but I don't think this is the time or place.

Participant:

Not lectures in general, id love to know it in the context of avalanche.

Participant:

I'm well versed in consensus in general.

Participant:

Are there any timelines that we can expect? For e.g by EOY 2019?

Kevin Sekniqi:

Big if true.

Kevin Sekniqi:

Yep!

Emin Gün Sirer:

We plan to provide a set of libraries to support common operations. And formal verification is a topic near and dear to my heart. We would love to work closely with any groups that have experience and toolkits to bring in.

Participant:

What is the word to describe the actual thing/network that Ava Labs is building? (I'm assuming the word "blockchain" is inaccurate.)

Emin Gün Sirer:

I spent a fair bit of time in 2017/18, going to the formal verification community and trying to get them interested in cryptocurrencies in general. I think these efforts paid off a bit, and we're seeing more engagement from those experts.

Kevin Sekniqi:

We've had trouble too! Not sure, it's not shards for sure, it's more like quasi-independent subnetworks with their own engines etc, but they can swap super quickly and freely.

Emin Gün Sirer:

Ava. It's a platform.

Moderator:

What is the Ava monetary policy?

Participant:

Does AVA have investors? And if so are their tokens vesting over time, is there a cliff, do they have any immediate access to liquidate tokens? I ask because this happened with Algorand where investors purchasing at 5 cents sold for a 20x shortly after the tokens became available to retail. This then lead to a negative effect on retail who lost money and confidence in the project (since this environment is so speculative in general) Are you planning to avoid this pitfall?

Moderator:

🕘 15-minute warning! Get your questions in.

Participant:

Is AVA trying to be money?

Emin Gün Sirer:

Great question. Ava enables stakers to mint coins according to a set minting rate. Unlike almost all other coins (which set these kinds of parameters in stone at the time the network is created), Ava leaves this rate to be determined by the network participants, within limits.

Kevin Sekniqi:

If we solve a real problem for the financial world, and not just provide meme CDOs, then we will be as real as USD.

Participant:

Roughly when do you expect the Ava mainnet to launch?

Kevin Sekniqi:

This is part of our governance mechanism, which is built-in. It's effectively dynamic re-parametrization.

Participant:

Key management across crypto as a whole is often seen as an issue.  Are there any differentiating factors that AVA plans to address here?

Emin Gün Sirer:

Yes, we have taken seed funding to get started. We are working very hard to avoid the pitfalls we have seen before us. Further, we feel that we have to do justice to the amazing level of decentralization that is only possible with our protocol. So we will work very hard to make sure that there are as many participants as possible to the network during the initial coin distribution.

Participant:

The investors have a lockup? If yes, any percentage is unlocked right away?

Emin Gün Sirer:

Yes! We plan to support "covenants" and "vaults." Vaults allow people to revert thefts that are caught within a certain time period, without impacting fungibility. They were invented by Moeser, Eyal and myself, and we plan to incorporate them into Ava.

Participant:

Will there be any numbers published on cap tables/vesting schedules before retail trading happens post-ICO?

Participant:

I assume you won't be able to include normal retail traders (non-accredited) for both logistic and probably legal reasons

Participant:

Is that correct?

Moderator:

Can you explain a bit more about how Ava governance works and the motivations for choosing that design?

Emin Gün Sirer:

Frankly, we do not know yet. How to bootstrap a PoS coin in a way that is compliant with regulatory requirements still remains, despite the entire industry's best efforts, a mystery. The SEC has provided no "bright line" guidelines for us. 

As a matter of principle, we are committed to carrying out a fair distribution of the tokens in a compliant manner.

Participant:

What are some of the negative trade offs that Ava chooses to take, if at all? For e.g Ava is worse at X but better at Y

Kevin Sekniqi:

Great question! I'll take this.

Participant:

Thank you. I hope you choose to publish the distribution numbers and vesting schedule (even in aggregate). We can assume all intelligent investors would behave the same way if there's a 10-20x to take off the table. Otherwise it's impossible for retail traders to make informed decisions.

Kevin Sekniqi:

As a general principle, we heavily optimize for product fit and then construct the best possible technology we can to fit that product. This is something that I sincerely have no seen in other crypto teams, and it's concerning for the space. 

On the technological end, we've decided to place significant emphasis on safety over liveness. We would much rather wake up one day and hear that the network has been down than hear that people have lost their funds. 33% is, in my opinion, far too low of a bar for other systems that are using it.

Emin Gün Sirer:

There are two overriding principles:

- We want the key parameters of the system to be able to change. Other coins, which typically fix these in stone, often end up missing their targets. E.g. ZCash seems to be over-rewarding its miners right now, and changing that is going to be very difficult for them. 

- We want the system to be predictable. Cryptocurrencies should be a calming, non-anxiety producing technology. 

The first principle motivated us to use the network as a crowd oracle, in essence, enabling them to say things like "hey, let's mint at a slower rate" or "let's mint faster and attract more stakers" and so on. They can change key parameters like minimum staking amount, minimum staking duration, and minting rate. 

The second one limits the changes. The system cannot suddenly vote to change the minting rate by any more than 10% of its current value. This way, people can go on vacation, not monitor the system, and come back to a predictable universe.

Moderator:

Gotcha, that's very helpful. Thanks emin

Moderator:

Alright folks we're up on time! Everyone, please give a huge thanks to the Ava Labs team for coming on today!

Moderator:

Kevin, Gün - what's the best way to stay up to date with project developments and get in touch?

Kevin Sekniqi:

Ya this was awesome! We have so much more to share but we also have some people taking notes 🙂. But in due time we shall share everything!

Emin Gün Sirer:

Thank you for having us @spencer This was a lot of fun.

Everyone, you're most welcome to our official Telegram Channel!

Emin Gün Sirer:

https://t.me/avacoin_official

Kevin Sekniqi:

We also have https://t.me/avalanchecoin

Kevin Sekniqi:

And of course twitter.com/avalabsofficial

Participant:

Thanks @kevin and team :) enjoy your weekend

Moderator:

Good stuff. Thanks again guys and thanks Crypto AMA for tuning in and asking great questions as always 🙂

Participant:

🙏

Participant:

Thanks @spencer

Participant:

Thank you @kevin and emin for your answers :).