Transparency within the Dfinity Foundation

[Leamsi]

I think the problem with this is the “because I feel”. That is laudable of you, and as it should be, but it shouldn’t be left to dfinity employees’ discretion. What if a colleague does not “feel” they should not benefit from a project they decide to fund? What if on the contrary they “feel” like tipping the scales to benefit a friend, relative, or independent business venture? Conversely, what if you discover accidentally that a project you approved had one of your relatives as an investor? How would you protect yourself against accusations of self-dealing, in the absence of transparent procedures you can point to and say “see? I did things impartially, and awarded them the same as all the others, following our published and vetted process”.

These things protect both, the community, and the people involved in awarding the money.

[Leamsi]

Is it? These are great tweets. Where are the processes? Where are the binding, institutional rules that would make non-compliance an enforceable breach? What if in 2 months’ time he changed his mind. Would he tweet about it? What if he absolutely follows through: only buys Dapps publicly. Declares an acquisition. Does this cover acquisitions by third parties in which he owns a stake? He only holds ICP in his own name. Does he hold 10 other currencies through a company? In a trust? In a relative’s name? Far more important, is there absolutely any rule in Dfinity that would make the above an enforceable breach? What is your own instinct?

Here is an example. In 2017, Dom was both the head of Dfinity, and the CTO of String Labs. Dfinity raised c.$4 million then. In other words, Dfinity paid Chief Scientist Dom to raise millions to pay CTO Dom to work on products to enhance Chief Scientist Dom’s Foundation. For a non-profit, it sounds pretty profitable for Tom. But it’s a bit murkier than that. According to Crunchbase, String Labs raised a total of $5 million in from 6 investors in a Seed round. That’s it. The round relied on Dfinity. But Dom’s String Lab co-founder, widely advertised that String Lab received $25 million from Dfinity Foundation. And then it closed.

So Dom paid Dom to get 6 investors to pay Dom $5 million, and then Dom paid Dom $20 million more from Dom’s non-profit foundation. Did Dom make a profit from Definity? I’ll let you be the judge. String Lab appears to be closed down. How did it spend $25 million? Definity god knows. And Dfinity god’s partner, Tom.

So forgive my scepticism if when Dom Tweets about his transparency, I parse his words extra carefully, and ask whether he intends to write down the same thing in anything more binding than a Tweet.

[toby2911]

Well made points!

Time and time again, it appears that Dfinity’s lack of transparency is what is most likely to kill the project. The lofty goals of ‘decentralising the Internet’ are almost laughable when the above is considered.

As mentioned, why would anyone choose to build on a platform that sells itself as ‘reducing platform risk’ when it arguably offers greater platform risk (project failure, lack of transparency, law suits, untested engineering and blind faith in Dom).

All very frustrating considering the amazing team and tech.

[quinto]

I’ve been watching this thread develop over time, and sure it gets more & more interesting. Although I applause you (@leamsi) for pointing out the elephant in the room and I agree transparency with the foundation is an important aspect for sure, I’m not convinced it is as detrimental to the project (i.e. Internet Computer) as you made it out to be.

As a developer myself, I am more interested in knowing when or whether IC can grow independent from the dfinity foundation. It is unavoidable in the early days that a platform needs its original creator, but is the tie between IC and the dfinity foundation grow stronger or looser?

On one hand I see accusations of the foundation “dumping” tokens (to which Dom have denied), on the other hand I see worries of dfinity having too much voting power in NNS (which effectively controls IC). But these two are conflicting opinions to each other, aren’t they? If anything I’d rather see the foundation diluting its ICP to investors sooner than later, so that voting power can grow more decentralized.

Voting power aside, the ability to continue develop and advance the technology behind IC is also very important. At the moment the leading role has been assumed by the foundation. Are they doing a good job? Are they incorporating or rejecting community contributions? Are they holding onto the expertise or are they willing to share? Is there any other entity (with vested interest in IC) that can do a better job than the foundation? If there is not, why not? How can the community foster a group of open source developers for IC core tech without depending on dfinity?

I think these are right questions to ask, instead of lingering on the transparency issue of a single entity: the foundation. I for one has little interest in knowing how dfinity manages its teams, how much control Dom exerts on his organization, let alone how much he gets paid. The mentality of relying on a single entity for IC to thrive must change, otherwise it never will.

That is of course not to say, from a legal & financial stand point, DFINITY should not improve its transparency, or the questions you ask are not worth pursuing. I think they are important, but for different reasons.

It has become a common & dominating practice for crypto startups to not to have “shareholders”. They effectively run token sales to raise money, but has no legal obligation to answer to tokenholders since they are not “shareholders”. Dfinity Foundation is no exception. Other foundations may do better at financial transparency, may do worse. I don’t really care if any of these foundations survive, what I care is whether the product they build will survive. A platform can only survive by being decentralized, both in how the platform functions day to day, and in how its future development is carried out.

I will be glad to see at some point the balance of core IC technology development tipping more towards the community. Are we on the right course? Does financial transparency help to stay on this course?

[Sabr]

@Leamsi

I genuinely love the concept of the Internet Computer. I hugely appreciate the extraordinary brain power, innovation, creativity and community. It is deeply seductive, because I want to believe. I am perhaps somewhat shielded by the lack of a personal profit motive: I’m not imagining myself an ICP millionaire in addition to this wonderful project.

Ditto on every point for myself.

@Leamsi

1 guy. In total, absolute control. Can shut down Dfinity tomorrow. Can replace anyone. Can veto any project. Or shape it. Can implement a backdoor (not saying there is one, just speaking to the power of his position). A single monarch of Dfinity, as an evangelist of the utopia of decentralisation. Cognitive dissonance anyone?

Ditto again for me. This is THE key sticking point for me in my due diligence before I can invest any significant time or money into the IC. The IC is fundamentally run by a dictatorship presenting itself as a democracy, and I don’t see any clear sign yet that they want to evolve away from that. That said, I don’t think it is necessary to ascribe any conspiratorial or deceptive intent behind this fact, since there are some significant theoretical hurdles to cross to achieve true decentralization. Also, many DAOs who have become mesmerized by the promise of Web3 make similar over-exuberant claims of being democratic when they are most certainly not.

As I’ve stated in several other posts, real power resides in prioritization, which the NNS is completely incapable of. Such decisions involve a prioritized choice among competing alternatives, particularly when the allocation of scarce resources is involved. NNS approval voting (like in all DAOs today) is almost always just a yes/no rubber stamp on a single isolated alternative/proposal, since it is deeply dependent upon the priorities that have already been decided upon by those with the real power to allocate resources.

@Leamsi

I don’t think it’s FUD (“heretic!”) to question, in a collective endeavour to build a decentralised internet why Definity (ahem Dom) choses to structure itself as an uber-centralised pyramid calling the shots with a single man at the top with absolute no accountability and zero checks and balances.

@Leamsi

In closing, I just want to emphasise the key word in my last two paragraphs: choice. Dom is not tied by some existential constraint. No FUD goons are pinning Dom to his chair. In one instant, he could add seats to the board. Include ICP investors, creators, advocates. Literally in one instant. He could tell his token Swiss lawyer: please copy and paste the statues of 99.99% of non-profits and reword for our purposes. Voilá. Checks, balances, accountability, and trust, for when faith won’t carry you.

Choice is most definitely at the root of the problem - in this case, collective choice. Collective prioritization is a far more difficult problem to address than almost everyone imagines it should be, as I’ve highlighted in previous posts. It is the core topic of my upcoming PhD project. I’ve even offered to DFINITY that we collaboratively explore ways to cross this “prioritization chasm” as part of my research to make the IC as truly democratic and decentralized as it presents itself to be. However, I have heard no response to that offer yet, made in both public posts and via DM. If they are not even interested in exploring how they can achieve true decentralization, then that would definitely be a big red flag for me. The jury is still out, though, so let’s not jump to any conclusions yet.

Again, I understand your frustration, @Leamsi, but I think it is a distraction to imply that the problem must be the particular dictator in power - benevolent or otherwise - rather than the system itself. No DAO that I’m aware of has adequately addressed this prioritization problem, let alone solved it, so try to keep that in mind. That said, I do agree that some mitigating measures around financial transparency and internal controls, with independent verification, would be very beneficial in the meantime.

[Manu]

hey everybody! I would like to share my personal view here, more as a community member than as a DFINITY employee.

If we look at other web3 projects, there are always companies or foundations contributing to them, and those companies or foundations are not truly decentralized. Some may be more transparent than others, but at the end of the day there is always some board that is in charge. However, what is important is that the project overall is decentralized. As the ICP community, I think we should focus more on decentralizing the IC instead of only focusing on DFINITY.

I believe the lowest hanging fruit is NNS voting: DFINITY has 23% of the voting power today, but since many neuron holders choose to follow DFINITY’s votes, DFINITY actually has an absolute majority on most NNS topics (if we include the follower votes). I think this unfortunate, because the IC’s NNS is actually a super powerful DAO, because all replica software and NNS canisters must be approved by the NNS before they can run, so the IC is truly governed by the NNS. I believe we need to make sure that the NNS vote really reflects what the community wants. That way, there can never be any change made to the IC that does not have the support of the majority of the community.

Some concrete steps we can take to decentralize the NNS vote better:

• implement “periodic confirmation”: right now we have a huge chunk of voting power (seemingly 50%) that follows DFINITY on “all topics except governance” but still hasn’t configured following on the governance topic, indicating that they’re not actively participating. This group follows DFINITY today, so periodic confirmation would likely help decentralize the vote quite a bit.
• encourage more community leaders to step up as voting neurons, and in particular encourage them to vote on more topics than just governance. I think very important topics are “system canister management” (which approves upgrades to NNS canisters) and “replica version management” (which approves new replica software versions, we recently split this off into a separate topic to make it easier for the community to vote on this topic). It is critical that the Internet Computer replicas and NNS canisters only run vetted code. Voting neurons can get the source code for the NNS canister and replica upgrades, ensure there are no unexpected changes in the source code or or git history, and build the artifact from source to ensure that the binary is indeed built from the claimed source code.
• encourage voting neurons to advocate their neuron more actively and to explain why they voted in a certain way
• encourage everybody to revisit whom they follow, and consider following community neurons instead of all following DFINITY.

[Leamsi]

I think NNS prioritisation is most certainly a crucial problem to solve. I have elsewhere pointed to the problems that stem from the huge concentrations of voting power. Liquid democracy is certainly liquid, but not democratic. Still, these are precisely the design challenges that make a project like this interesting and exciting: assuming good faith and capacity for influence. You have just pointed to your own need to collaborate with Dfinity to take your doctoral ideas for improving ICP forward. There is no need for DFINITY to take on your idea: they might have different priorities; different implementation philosophies; different people already working on it. But the fact that you would see a need to collaborate with DFINITY demonstrates the centrality of DFINITY in the ecosystem, even after renouncing a range of votes. And DFINITY is Dom. So Dom’s ethics, choices, priorities and motivations will cascade down to systemwide effects. This makes all your NNS development efforts, even if successful, a hostage to fortune. I think we are a VERY long way from a scenario where Dfinity is just one more voice among equals. While the absolute executive, financial and technological control of Definity resting in one man, with not an exactly uncompromised track record, then, as I put it, all other development rests upon a house of cards. Dom’s cards.

[Zane]

Afaik IC’s code is not completely FOSS, so if Dfinity ever went rogue the community can’t easily fork the protocol, that alone should impose a higher level of transparency on Dfinity’s end. The project relies on its good faith way more than others.

[passio]

A good dictactorship is better thank a false democracy. Dfinity is made of most brillant mind in the world, probably with some of them working to change the world more than because the need of money. For this reason, I am not bothered by their large power on the network. The perfect democracy can’t exist in this world… I saw many people and project who tried… if it could work, the Bitcoin, appeared in 2009 would not be what it is today. I realized now that it’s better to have trusted human guarantors, even if it’s very imperfect. If Dfinity didn’t have all this power, I would have jumped ship.

[Leamsi]

Excellent and thoughtful answer. I couldn’t agree more: if DFINITY was incidental to ICP, what Dom does or doesn’t do with the money he made legally or illegally, ethically or unethically (not the same thing), lose any interest to me. And yes, the question you ask, given the answer you hope for, would resolve the biggest vulnerability I see in the project, replacing it with other, design focused concerns.

“As a developer myself, I am more interested in knowing when or whether IC can grow independent from the dfinity foundation. It is unavoidable in the early days that a platform needs its original creator, but is the tie between IC and the dfinity foundation grow stronger or looser?”

Those are two different questions. The answer to the second is clearly that yes, the relationship is growing looser, given DFINITY’s laudable choice to abstain on a range of strategic NNS votes.

But in terms of the systemic vulnerability to Dfinity’s dangerously awful governance standards as a (nominally) non-profit foundation, the question is not whether it’s looser or stronger, but whether it is decoupled enough for Dom’s total control not to constrain, shape or endanger the unfolding of IC. And I think the very fact this thread is unfolding, shows that DFINITY remains very, very far from peripheral to IC. If Dom sneezes, there is nothing approaching the social distancing required to stop IC catching Covid.

And yes, I hear the messaging from DFINITY about wanting to step back, acknowledge the positive steps in this direction, and I’m even predisposed to believe in the intentions and good will of the team as a whole. But: is Dom’s position, legal ensconcement, doubling down (“advising”?) on lack of disclosure, and current activity online the sign of someone wanting to step back, chill out, and watch his baby leave the nest? Or do those indicators suggest someone still deeply invested, personally and economically, and ready to make decisions to keep a steering hand quietly hovering near the scales?

And if the latter, does the current direction of travel of DFINITY allow you, as a developer, to build confident in the direction of travel, when Dom, and Dom alone and exclusively could, tomorrow, decide that Dfinity resumes its NNS voting? Dom and Dom alone, once everyone has built good infrastructure on the premise of DFINITY receding in control, could go, thanks for all the fish, on second thoughts we’re back? Or absolutely any other scenario that comes to him?

I think that Dom would be delighted with the community’s “ability to continue develop and advance the technology behind IC.” Free labour on a project he controls, and where any retreats are a sign of his largesse, and subject to his fiat.

I agree the questions you ask are the most critical ones. I think we are a long way from being able to answer them in a way that makes the sole control of Dom over dfinity with zero disclosure, accountability or transparency, anything other than IC’s key constraint and biggest risk.

[Sabr]

@Leamsi

There is no need for DFINITY to take on your idea: they might have different priorities; different implementation philosophies; different people already working on it. But the fact that you would see a need to collaborate with DFINITY demonstrates the centrality of DFINITY in the ecosystem, even after renouncing a range of votes.

There is actually a dire need to incorporate collective prioritization into the NNS, and into every DAO, since there is no way possible for power to be decentralized without this capability. However, if their hidden intention is retain their centralization of power indefinitely, then I agree that there would technically be no “need” for them to explore any decentralization options. In fact, they would actively oppose such efforts. Hence my due diligence test of their intentions.

As for “different people already working on it”, I have done a fair amount of review of the literature and potential individuals at DFINITY with relevant backgrounds to address this problem. I am fairly certain that no one there is even seeing this as a major problem yet, let alone working on how to solve it. More likely, they have no idea how deep this academic rabbit hole goes and how many academic fields it cuts across.

[Leamsi]

Sorry, “need” was clumsy. I meant obligation. DFINITY is not obliged to adopt your idea. The point I was making was that DFINITY remains so inseparable from IC that you feel a need for their buy-in (Dom’s buy-in) to take that agenda forward. Is there a reason the current proposal system is not a sufficient channel for you to have your vision adopted into the system?

I would be very interested to read your research, btw. I think the oligopolist effect of liquid democracy in every crypto I’ve looked at is the biggest systemic challenge to web3 and I get a sense your PhD attempts to somehow keep the status quo in coin distribution yet find a way to make decision making not determined by coin concentration? Anyway, discussion on this specific subject of prioritisation belongs on another thread. If you point me one or start one, I will be happy to continue the conversation there.

[passio]

The DFNITY’s power on the network is not the problem, it’s the solution. As Bitcoin high cost in energy is not the problem, but the solution. If DFINITY would not have this power, it could be very very very dangerous, and more dangerous at this low price.

Their power will diminish a bit over time, but to begin with, their power must remain strong to protect the network and make the best decisions. There would be powerful actors, who would have interest to destroy the ICP network. The DEFI on ICP is a threat for all other blockchains… Some may want to eliminate this threat. They are ever trying by dropping the price, but If they accumulated too many tokens, they could influence the votes to make terrible decisions.

From my point of view, those who do not understand what I am saying, can only be stupid or enemies of ICP.

DFINITY is made of smart and good people, if you don’t let them keep a lot of power, of people will lot of money will take it.
There are no debates. There are no debates. There are no debates.

If you don’t understand this, you are crazy, and you probably understand absolutely nothing about how the world works, and nothing about how cryptocurrencies markets works… And you’re a perfect example of why the democracy is more dangerous in hands like yours rather than in the good hands of DFINITY dictators.

[Leamsi]

IC’s code is not completely FOSS, so if Dfinity ever went rogue the community can’t easily fork the protocol

This is an absolutely, non-negotiably critical point. The entire IC boils down entirely to the undisclosed, un-reportable and unchecked decisions of just one man. Who oversaw an ethically compromised Genesis. You’d think, at the time when SBF, the genius, the philanthropist, the progressive, brought down FTX, IC people would be a little more cautious. What do so many of the spectacular collapses in crypto have in common? Sole power and zero transparency concentrated in a single CEO. Again, you’d think by now that transparency, accountability mechanisms, and ideally power sharing, would be one of the design principles that would be baked into newer initiatives, for their own wellbeing.

[Sabr]

@passio

A good dictactorship is better thank a false democracy. Dfinity is made of most brillant mind in the world, probably with some of them working to change the world more than because the need of money. For this reason, I am not bothered by their large power on the network. The perfect democracy can’t exist in this world…

I have commented along these lines too in another thread, per my excerpt below. In short, a dictatorship model can be effective and lower risk in the startup phase, but it will not be at all appropriate in any scalability scenario that is remotely close to what DFINITY envisions for the IC in the future. That’s why we need to start exploring collective prioritization options for the NNS now, since that capability is totally nonexistent now.

@Sabr

The Internet Computer won’t be perceived as a beta project for much longer. As the stakes become higher and more real, the cracks in the decentralization narrative will become exponentially more obvious and repugnant to those who depend on the IC to sustain their livelihoods and ecosystems.

In the meantime, we just have to accept the following: 1) DFinity is a single point of failure in this visionary project, which (if unchanged) will increasingly endanger its fundamental mission to oppose both centralized power and censorship; and 2) DFinity is essentially run under a traditional dictatorship model, like every organization on the planet today that is too large for deliberative democracy and other consensus reaching processes.

From my perspective, I am quite OK with this concentrated risk and dictatorial power during the startup phase. It could even add to both innovation agility and the prospects of success given the trust that I currently have for the DFinity team and its vision. However, in the longer term, I think we all agree that we should never have to rely on trust, and we most certainly can’t rely on any dictatorship remaining benevolent or censorship-free when the economic and societal stakes for the IC grow exponentially higher. That is why I have been so emphatic about initiating some sort of collective effort to cross the “prioritization chasm”. This is the only way the decentralization vision can become a genuine reality – in bulletproof fashion – instead of just being an illusion of decentralized power that we have today.

[Leamsi]

The DFNITY’s power on the network is not the problem, it’s the solution. As Bitcoin high cost in energy is not the problem, but the solution. If DFINITY would not have this power, it could be very very very dangerous, and more dangerous at this low price.

Fine: let’s go with this, in my view, somewhat naive perspective. Let’s accept the plebs cannot be trusted, that decentralisation is not really that big a part of the blockchain vision, that our betters know best. I accept: let’s keep the balance of power, not in DFINITY, but in Dom as the sole controller of DFINITY.

Now tell me the many advantages to the system of that absolute power requiring no transparency, no disclosures, no reporting, and having no procedural, enforceable boundaries to ensure fair play. I’ll wait.

The problem is not authority: it is accountability and transparency: whether it is by Dom, by DFINITY or the NNS voters. So even if we go with your hyper centralised utopia, we need to make the exercise of that absolute power transparent, rational, beneficial and accountable. Even if we’re not worthy of holding the power, we are surely worthy of knowing how we can avoid its abuse, and if abused, get some form of redress and correction. Or are we not worthy, even of that?

[Leamsi]

@passio

A good dictactorship is better thank a false democracy.

So what’s the difference between a “good” and “bad” dictatorship? I would suggest a good dictatorship is one that is transparent and accountable. A bad one is one that is neither. A good dictatorship asks you for trust and gives you the elements to verify. A bad one, only gives you the option of faith, while holding every single thing you build, invest or purchase in IC on his sole, unfettered, unrestricted hands. Is that really your optimal scenario?

[Sabr]

@Leamsi, feel free to continue this conversation in the governance thread, where I first brought up this topic in my very first post on this forum 23 days ago. Check for my follow-up posts there too.

Technical Working Group: Governance

My Introduction and Governance Interests Hello everyone. This is my very first post on the DFinity forum. I’ve done my best to find the appropriate place for this post, so please redirect me if I didn’t hit the right category and thread. I am quite interested in contributing and collaborating to improve NNS/SNS governance, so I would like to be a part of this technical working group. I am not a software developer, but I have other technical skill sets. I am currently scoping out an experimenta…

I agree with your view that DFINITY is currently inseparable from the IC, hence the key problem.

Approval voting per the NNS and every other DAO has no prioritization capability amongst competing alternatives, nor will it ever have this capability due to its fundamental lack of information granularity. All the toughest and most important decisions generally involve prioritizations and tradeoffs between alternatives, not just a yes/no approval on isolated alternatives/proposals.

The problems with liquid democracy as currently implemented are actually much easier to fix. However, it doesn’t make a lot of sense to try to fix this problem without implementing a collective prioritization capability first. For example, the latter can actually be used to fix the voting power problem simply by allocating voting power based upon a vote on the evaluators themselves (i.e., where evaluators or classes of evaluators are the alternatives being collectively prioritized to determine their respective voting power/weights).

[kyliux]

@Leamsi

at’s the difference between a “good” and “bad” dictatorship? I would suggest a good dictatorship is one that is transparent and accountable. A bad one is one that is neither. A good dictatorship asks you for trust and gives you the elements to verify. A bad one, only gives you the option of faith, while holding every single thing you build, invest or purchase in IC on his sole, unfe

Dictatorship is something necessary for short term changes and big step in evolution. It is fair that Dfinity used this for safety reasons during the decentralisation process.

I am personally convinced that Dfinitys intention are to go in the right direction and give gradually more power to the community.

Yet I have few concerns to formulate :

1. For the community to make good decisions, we need to be informed and also educated. We need a open source book of the specification of the functionment of this non profit association, its finance and directionm priority, stakes etc…
2. This information has to be provided progressively ( we cant probably swallow all stakes in one shot )
3. This information has to be delivered in a good Timing. I guess if we do not have this information yet, its because the Bird cant really fly safely yet.

The big cons I see from Dictatorship, is that if you delete the leader, the whole house of cards likely collapes. And imo Doms safety alone is a good enough reason to decentralize this decision weigh as soon as possible. I am also convinced that he knows this and that it is on his best interest. So I think he will do it as soon as he can.

If someone would just confirm my hypothesis and tell me that * the time didnt come yet * I would understand.

But then, when ?

cheers

[Sabr]

@Leamsi

The problem is not authority: it is accountability and transparency: whether it is by Dom, by DFINITY or the NNS voters. So even if we go with your hyper centralised utopia, we need to make the exercise of that absolute power transparent, rational, beneficial and accountable.

Agreed. To follow upon that argument, let’s assume that Dom has been somehow verified to be a saint. Ultimately that doesn’t decrease the risk at all. Dom could get hit by the proverbial bus any day. And one day, with absolute certainty, a new unknown dictator will take his place. That’s why I find it a bit distracting to focus on one particular person here.