Geeq Founder and Chief Economist, John P. Conley, was on call to answer questions from the community. Questions ranged from the practical to theoretical. Thanks to everyone who participated, and to Qudem for organizing and hosting the AMA Live on their platform.
Here is a recap, in case you missed it. A few answers have been edited to include the most up-to-date information (via links).
GeeqChain is Blockchain V3.0. There are many blockchain 4.0 projects on the market. Isn’t it a bit late to say that now? Would you be able to be competitive?
I think in an earlier version of the white paper we referred to Geeq as Blockchain 3.0. By this we meant that we were building on the foundation that Ethereum (blockchain 2.0) established to advance DLT to a new level.
The most important element of Geeq is the Proof of Honesty validation protocol that provides 99% BFT and uniquely implements honest node behavior in coalition-proof equilibrium. In addition, Geeq’s interoperable multi-chain architecture creates an ecosystem in which different instances of geeqchain can be customized to suit the needs of different use cases and DApps with high transaction throughput at very low cost. I don’t know how I would characterize this in terms of blockchain versions, but I think it makes us very competitive with existing projects and likely to remain so in the future.
Hi. When Geeq will launch a public sale? Because I am watching this project since July 2018. Geeq seems pretty promising and I want to know will we be able to participate in the ico or ieo?
I really appreciate that you found us so early and have been following us ever since. I hope you have or will join our social channels : our Telegram has moved from its original group to https://t.me/GeeqOfficial and our Twitter is @GeeqOfficial.
From the beginning, Geeq has been very careful to be compliant about fundraising methods, so we have always concentrated on private fundraising efforts. I’m sorry, but for compliance reasons, I am unable to comment on any type of public funding. Please just keep checking our social channels for any announcements on the subject.
Hello, John. What exact functions to be carried out by your token in the GeeqChain?
The main purpose of the token is to pay nodes for their validation and virtual machine services as well as providing a means for nodes to stake a bond to assure honest behavior. $GEEQ is a general purpose transactional currency that is well adapted for micropayments, streaming payments, and any sort of value transfer.
Section 6 of the new white paper goes into a lot more detail of how $GEEQ could be used and why we think that the platform does an especially good job for these use cases.
Can you tell me if any venture capitals have invested in this project? And when can we see the Geeq on exchanges?
Thanks for a very interesting question! The short answer is no, Geeq has not had any venture capital funding. From a fairly early point, we realized that we had a small, extremely dedicated core team who were all on the same page about what Geeq was doing differently and what it could become. The best advice my supervisor gave me in graduate school was that you could either read or write. For Geeq, this translates in a choice of market and promote, or develop and build.
So, we had a small private sale of equity which helped us bootstrap the company ourselves, rather than spend a lot of time and effort on fundraising before we had anything. Any extra time and energy we’ve had has gone into outreach and education, and meeting with and getting feedback from potential customers.
We’re now at a point where we’re positioned to grow quickly, though, and we’ve posted our new Investor Deck on our website for anyone who is interested in our private sale.
As for when $GEEQ will be on exchanges, all I can say is that, due to compliance reasons, we cannot comment on exchanges, or when, or if there will be an IEO. I apologize for that very unsatisfying answer. Please keep checking our social media channels for any future announcements on this subject.
How much can tps geeq provide in mainnet? According to your whitepaper, geeq transaction cost will be 0.0001$ How does geeq serve high tps at such low cost? Which type of dapp is supported by geeq blockchain?
This question gets to the heart of how Geeq is really different.
First, we estimate a single instance of geeqchain (See Section 3.3 of the White Paper) can execute about 400 TPS using an ordinary residential broadband connection. Bandwidth is the bottleneck. If Geeq nodes were placed on AWS and given as much bandwidth as they wanted, TPS would go up considerably.
Second, Geeq is a true multi-chain platform. Any number of interoperable instances can be created. This means the mainnet as an ecosystem can collectively execute an unbounded number of TPS.
Third, Geeq is inexpensive precisely because of the its multi-chain architecture. Each instance of geeqchain has its oven separate network of validating nodes. Each network might contain between ten and one hundred nodes. Geeq does not rely on Proof of Work, so the only resources required are the compute, storage, and bandwidth used by ten or twenty computers to process, store, and communicate transactions. More on this below, but a Geeq node processing 20 TPS should be able to run as a background process on a laptop. The small size of the network and the efficiently of the message and communications protocol simply does require very much computational effort.
Fourth, each instance of a geeqchain has a separate application and validation layer. The application layer will initially be built to support Solidity, but other code bases will be supported in the future.
I wanted to know your token economics, will any tokens be allocated for bounty or airdrop.
Thanks for your question, We definitely would like to involve the community and crowd-source some of its wisdom. I can’t promise, because our legal team keeps a close eye on how regulations change and we are committed to best in class compliance. I know from the development side, however, we’re interested in bug bounties so we are actively exploring all possibilities.
As for airdrops, Geeq definitely has tokens allocated for airdrops for people who become active in Geeq! This AMA with Qudem is our first partnership involving an airdrop, so thank you for participating!
Will there be any token sale stages or directly IEO will be performed?
We will be holding a private token sale and our updated Investor Deck has been posted on our website.
Other than that, for compliance reasons, I cannot comment on exchanges, or when, or if there will be an IEO. Thanks for your interest and please keep checking our social media channels for any future announcements on this subject.
What were the major issues in modelling POH given it’s based on Game Theory and the assumptions behind it?
That is a pretty subtle question. My background is game theory and mathematical economics. Much of my work is in axiomatic approaches to bargaining and operations research, mechanism design, and implementation of non-cooperative solutions. I say this by way of apology for the nerd-sense that follows.
First, I assume that all agents are self-interested. There is no such thing as altruism.
Second, I assume I can’t detect Sybils. I have no way of knowing how many independent agents are behind the nodes in the validation networks.
Third, I assume that all nodes can costlessly communicate and collude.
Suppose I was able to prove that honest behavior was a Nash equilibrium? (This is what white papers actually mean when they claim that good behavior is incentive compatible.) What good is that? In almost all cases there are many other Nash equilibria, usually an infinity of them. For example, if all Bitcoin nodes suddenly decided to raise mining rewards to 50 bitcoin tomorrow, no single miner would want to defect. Any single miner that rejected blocks with 50 bitcoin rewards would find that he was only writing to an orphaned shorter chain.
One thing this tells us is that we need a consensus mechanism that has a unique equilibrium. If there are multiple stable equilibria, there is no foundation to think we will get lucky and end up at one we happen to like. The other thing this tells us is that Nash equilibrium is FAR too weak for an environment where there are mining pools, Sybils, and other forms of collusion. Nash is only proof against unilateral deviation by a single agent acting in isolation. Nash equilibria are NOT proof against deviations and manipulations by coalitions of agents.
So what do we need? Wait for it…. Coalition-proof equilibrium! In fact, we need honest behavior to be the only coalition-proof equilibrium. This is part of the security guarantee offered by Proof of Honesty.
PoH goes further and considers the meta-game in which agents might receive rewards for bad behavior that go beyond those provided by the consensus mechanism. For example, North Korea, the NSA, or anarchist hackers might simply want to burn the world down. If the American economy becomes dependent on blockchain (suppose the NYSE is moved to Ethereum), then the value of an economic attack might far outweigh any rewards or punishments associated with the consensus mechanism itself. This is a rabbit hole for another day, however.
What discussions have you had with industry to evaluate the adoption of Geeq going forward and have you had any expressions of interest?
While we are pretty focused on development right now, from the very beginning, we have known we would follow a model of developing Geeq so it would be immediately applicable to real world use cases.
Our CEO, Ric Asselstine, has had experience in developing and bringing to market technological full-stack solutions that are agnostic and yet highly adaptable. Our entire team is completely united behind the goal of building technology that fundamentally works to support blockchain solutions where they are appropriate. One of the ways to ensure solutions work is to work to customer specifications.
We have been very fortunate to date to have attracted commercial interest and to have discussions already underway to inform product and solution development. I’m not at liberty to disclose those details, except that the feedback has been invaluable and there is a lot of experience on our team who have been working to establish a sales pipeline.
So for how much time are you thinking to keep your private sale up?
We started making our research known as The Geeq Project in early 2018, and some ico sites added us to their platforms. We are aware there is some out of date information on some of those sites, as we have adjusted our tokenomics since then, and we will be contacting them with updated information shortly.
(Edited: Please see the Tokenomics page for up to date information.)
Can you explain how Proof of Honestly protocol (POH) works? Which types of significant problems geeq plan to solve in 2020? Will geeq launch it’s public sale on ieo or ICO what’s your plan on it?
The fundamental idea of PoH is that the Geeq protocol (and all other well designed DLT protocols) is deterministic. Given a set of candidate transactions and a ledger state, there is only a set of valid transactions, one block that can be built, and one way to update the ledger. Anything else is dishonest.
In most protocols (PoW, PoS, PoA, dPoS, etc.), nodes attempt to come to a consensus view on the correct block and the correct ledger state. The problem is that if the weighted majority of the stake, hashing power, authority, etc. decide that a block is correct, the protocol accepts it. Even if the block has bad transactions that are visible and provabley incorrect, users don’t really have any recourse but to accept the consensus view or walk away from their accounts. In other words, if the majority says that a cat is a dog, from a protocol standpoint, that is the last word. This fact can also lead to hard and soft forks and ledger reorganizations that violate protocol.
PoH leverages the determinism of its protocol in two ways. First, Geeq’s workflow assures that all the nodes in a given validation network consider the same set of candidate transactions. Each node then separately and independently decides which transactions are correct and what the next block and ledger state are. Geeq has no leaders or block proposers. All nodes are equal. Nodes share their view of the ledger state as they consider the next set of candidate transactions. The key is that either there is unanimous agreement, or some nodes are dishonest. Honest nodes are able to prove the dishonesty of nodes with a different view and can write audits to remove them from the validation network. These audits result in the confiscation of Good Behavior Bonds posted by the dishonest nodes to be shared among the honest nodes. This audit mechanism is constructed in such a way honest behavior by all nodes is the only coalition-proof equilibrium (which is a much stronger statement than honest behavior is one of many Nash equilibria).
You are probably asking what happens if only a minority of nodes are honest? Suppose there is only one honest node. How would this node enforce his audit against the rest of the network? The second major element of PoH is to let users, rather than nodes, be the ultimate arbiters of truth and correctness. Users (through their user client software) ask nodes to prove their honesty before they send their transactions. An honest node (even if it is the only honest node in the network) can prove itself to users. Dishonest nodes cannot. Since any agent can tell whether a node is honest, why would any rational user ever accept tokens on a dishonest ledger if an honest one exists? Such tokens are likely to have been stolen, and are likely to be stolen from any user who accepts them. Thus, users who allow their clients to work in their interests will never be fooled into accepting payments on dishonest, and therefore, invalid ledgers. As a result, dishonest ledgers and the tokens they contain are valueless. Attempting to transact on a provably invalid ledger would be like trying to pass off a drawing of a $100 bill scribbled on a cocktail napkin as the real thing.
Did you partner with Morpheus Labs only for your PoH protocol or was there something else with it. Also, what both parties have decided to offer each other.
This is a strategic partnership that allows both companies the opportunity to explore mass scale up. We share a thirst for blockchain adoption that we believe Geeq capabilities unleash. Morpheus has an established footprint and stellar reputation in this space that we believe will be an exciting complement with Geeq’s strengths in offering a strategically provably secure and extremely flexible blockchain environment, as well as Geeq’s truly unique strength, which is to unleash limitless, secure micro transactions.
If you are interested in the security aspects of Geeq, they are summarized in the White Paper in Section 5, and more detailed proofs and discussion are in our Technical Paper. We talk about Geeq’s ability to open markets for microtransactions in the White Paper in Section 6-2.
How long do you think it will be before quantum computing breaks existing encryption?
How soon will it be possible to break AES-256 with quantum is one way to ask the question. I was lucky enough to meet with Michele Mosca who heads the Institute for Quantum Computing in Waterloo (Geeq’s hometown) last year. His view is that this is probably five to ten years out.
Another way to ask the question is how soon will it be economically feasible to routinely break encryption of any given type. Here, I think we have a much longer horizon. Qubits are very expensive and require elaborate refrigeration setups to cool them to close to absolute zero. Even if a quantum computer was constructed that could break encryption, it would cost many tens of millions of dollars (maybe more). As a result, there would not be enough quantum computational capacity built to be routinely used to decrypt the billions of encrypted messages sent over the internet every day.
This means that in five or ten years, you won’t know for sure if your message traffic has been read or that your bitcoins will be stolen, but it will be unlikely for many years to come.
Finally, the good news is that there are quantum resistant ways of encrypting data and signing transactions. The cost is that either you must use a very large encryption key (which takes up valuable space in the ledger and on the network) or use a lot more computational effort. Geeq has an upgrade path that can incorporate these quantum resistant encryption strategies if and when they become necessary.
How much Geeq will you need to run a node?
The short answer is that we anticipate that the good behavior bonds needed will be on the order of $200 worth of $GEEQ. This may be higher or lower depending on what any given instance of a geeqchain is doing (high or low value transactions), how big the validating network is, and how much excess demand or supply there is of nodes wishing to join the networks.
1. How does Geeq face the problems of hard forks ? 2. Misbehaving nodes are ejected from the validation network and Good Behavior Bonds (GBBs), previously posted in order to join the network, are forfeited and used to pay rewards to the agents who called for the audit. Can you shed some light on GBBs, I can’t seem to find this explained in the whitepaper.
My apologies for this. I think this is all much clearer now in the Technical Paper, Section 4.2.
Geeq does not have hard forks. Code is law, and protocol rules cannot be changed nor ledgers reorganized. If a node attempts to do so, it is per se behaving dishonestly, such dishonesty is detected, and the node is ejected by the network and the GBB confiscated.
What have you learned from early versions of the product or service? What are some of the major roadblocks along the way?
I have to say that we have learned a lot along the way. We started in 2017 and spent most of our early effort thinking about protocol design. I was lucky enough to be spending a sabbatical year as a visiting researcher at Microsoft Research in Redmond, and was surrounded by hundreds of Ph.Ds in all kinds of computer science, cryptography, systems engineering, information science, and almost anything else you could possibly want. There were also people from the business, marketing, legal, and development sides.
I was probably a slow learner, but I came to realize that game theory, economics, and mechanism design, my specialties, were only part of the picture. For example, network systems and communications protocols create their own set of security vulnerabilities that are completely independent of consensus mechanisms and improperly aligned incentives. Many, in fact most, use cases can be addressed with a good database. Only some need blockchain, and even fewer of those needed a native currency, and when a cryptocurrency was needed, a poorly designed monetary policy could easily make an otherwise excellent blockchain project or platform fail.
It is really interesting to dive into technical and mathematical issues in CS, cryptography, biometrics, graph and network theory. What really struck home in the end was that the business side of crypto can’t be an afterthought. Blockchain is a very powerful tool, but when it is slapped onto a business problem where it does not belong, it is literally worse than useless. It adds cost and complexity and often decreases security. So, what I personally, and the team as whole, have learned as we have built Geeq is that we have to take an integrated approach from math to marketing. A weak link anywhere, be it legal, cryptographic, or building for a use case that isn’t really there, can make the effort spent on the rest of the project irrelevant.
Is there any interesting story behind of the birth of Geeq? I heard it is created by connecting between Ric and John, but why is Geeq?
I don’t know if it is interesting, but the shortish version is this. A friend at Microsoft mentioned blockchain to me and so I started to read about it. At first, I was thinking only as an economist so I wondered about ICOs and what might be a sensible way to value cryptocurrencies as distinct from equities sold in traditional IPOs. There were some similarities, but a lot more differences. Tokens, in almost all cases, are not securities. I ended up writing a short white paper and giving a talk in a blockchain conference in Houston in April of 2017.
Meanwhile, Ric, who has been working in IoT since before it was called IoT, was at a different conference. He was talking to someone about his full stack solution for connected devices and was asked when he planned to add in a blockchain layer. This was the first he had heard of it, and so he dived in to learn what he could. He found my paper somehow and just called me up to ask about it. This was in August 2017.
We started to work together on this idea, but I kept realizing that it would be difficult to build on existing platforms. I told him about those problems and said I had an idea for a different approach to consensus, and we quickly dropped the original project to focus on what eventually became Geeq.
When we started, I thought I had it all figured out. When I began to develop the protocol formally, however, I realized that at best, I had an interesting idea for a mechanism. The protocol had to live in a technical ecosystem of networks, hardware, cryptography, and so on, and integrate all the constraints that followed in their trails. It was a lot harder than I had imagined. I also had very little understanding of what it meant to bring a startup to life, to support it with legal, accounting, and other administration services, not to mention the minor problem of finding customers. Fortunately, Ric had experience in all these things. In the end, we each learned a great deal from the the other and we all lived happily ever after.
Which are the key points of (PoH) that you see are outstanding compared with PoW & PoS or BTC and ETH?
The main points are that PoH is 99% BFT and uniquely implements honest behavior for nodes in coalition-proof equilibrium. In addition, we don’t require wasteful and expensive hashing races like BTC, ETH and other PoW approaches. Finally, PoH can’t be subverted by wealthy agents who can either afford to deploy large number of ASICs to gain a majority of the hashing power or who can buy a sufficient fraction of the staking tokens to control the chain.
Thank you to everyone who submitted questions!