[talk] the age of networks and the rebirth of cybernetics

The mainstage one from web3 isn’t posted yet. It seems like they are rolling them out one or two at a time and sharing them on social media. When I see it, I’ll report it here.

1 Like



Good talk. A few lines I jotted down that I think are relevant to SC:

in governance systems, “Can’t have high attention cost to participating

this. also enough skin-in-the-game for your vote to count.

we’re building a service infrastructure - providing decentralized services to each other

interesting framing. having done a bit of freelancing, and worked in OSS DAOs the last year, i kind feel like this reality is here, but it’s still very similar to traditional contracting.

bonding curves are better estimators of price/value….uniswap can serve as better estimator as price…

still don’t quite understand bonding curves (they sound suspiciously complex). but i do see an explosion of tokens, everyone/org having their own money. and making markets is a difficult, capitol and labor intensive process now. uniswap, bancor and other ‘liquidity service providers’ that automate this and lower the barrier to near zero are game changers, somehow…

any time we put an objective function in our systems, we make a subjective choice

something to keep in mind as we pick objective functions, as we’re actively doing now


Bonding Curves are about as complex as high school physics math. The problem is that unlike high school physics where there is one set of energy conservation laws in nature that all physical systems more or less abide by, the “bonding curve” style logics allow you to construct your own invariants. In fact whether we think of it this way or not, some sort of conservation law will likely get embedded in the mana vs cred vs money conversion equations.

In my talk I mention that uniswap is a proof in life of such a system. Here is some material (using data from the ETH/DAI uniswap and a cadCAD model of the smart contract mechanisms) from @markusbkoch who you met in a previous SourceCred office hours:

Video Demo: https://lnkd.in/gPhaM5P

Community Forum: https://lnkd.in/gFMVqHG

Notebook Viewer: https://lnkd.in/g4msAas

Tweetstorm: https://lnkd.in/gDUrn5s

Personally, I believe that constructing custom “energy conservation” type laws is how we will make messy web of composable smart contracts even remotely interpretable. Like so many things they start of complex, but as we come to understand them, I expect people to be able to put them to good use in both design and analysis of crypto-economic systems.

We (my co-author was @zixuanzh who you have also met) mapped out the mathematical framework for this kind of thing in late 2017, and published it in early 2018. Since then I’ve seen a steady rise in designs which can be explained in terms of these kinds of Lyapunov-like “value functions”, even if people don’t realize it. That said, its a cross-application of methods from systems theory, so I don’t expect to get particular mainstream. At least not any time soon.

Let also add that I am not really a proponent of tokens as “money” but i do see the explosion of tokens as a correction to the reduction of ‘value’ to one dimension of money. I think highly heterogenous tokens (in the back end of systems) will help our technological systems respect the fact that value is a highly contextual concept, and requires more degrees of freedom than merely say “something is worth a different of money to different people at different times”. I am very curious to see how the tensions between the need for more degrees of freedom and the limits on people’s attention will play out. I suspect we’ll see people getting involved closely with some ‘communities’ and caring to participate in processes that require attention to something tokenized (even if they don’t see the concept of tokens directly) but for vast majority of the things they interact with they will not.

Sometime I should write a whole post just on how I see individual versus community level self-actualization playing out. In my opinion we need both, and while they may be in tension, the ability for a disagreeing individual to choose between fighting for what they believe an organization should do and leaving to find an organization that fits their value system better plays a big role in it. I think there is a lot to explore here, and I actually think it also becomes relevant to SourceCred as it relates to the interplay between individuals and collectives.


Moderation note: I spun this out of the CredSperiment Progress Report since it stands alone well as its own topic. :slight_smile:

1 Like

Whoa…bonding curves :open_mouth: That’s gonna get interesting. Totally gonna work generally I think, but the outcomes are gonna have some unexpected dynamics.

Have also been thinking about tokens as multidimensional. While the utility token idea seems to have stalled, I think there is definitely value in embedding different “values” or “contexts” into money. Perhaps not because it makes tokens more convenient to transact with (they might always be a PITA to transact with), but because it allows for more expressiveness in transactions. A lot is lost in the “reduction of ‘value’ to one dimension” (indeed, the flattening effect of commodification generally is something I’ve become acutely aware of these days). Perhaps these tokens evolve far enough away from traditional money that they need to be called something else. If anything just as a “rebranding” because money has so much baggage.

Already I see people in crypto communities spending a lot of effort discovering and enforcing a community’s values/politics. It’s even being taken seriously as a necessary feature of the currency (e.g. Bitcoin maximalism). Not to mention general tribalism (I occasionally go to war with Dash shills on Twitter, which is the closest I’ve every come to being a sports fan yelling at a TV).

The individual/collective thing has been pointed out as one of the main drivers of various civilizations for some time, and I tend to agree. I think what’s fascinating in crypto is that a lot of the various relationships/power dynamics become explicit, embedded in the infrastructure. I am for sure in crypto because I see it as the best path to self-actualization (my main goal). I am also witnessing the tensions you speak of play out. As a general tool, I think SourceCred will be applied very differently in communities with different priorities regarding the individual vs. the collective. Or at least that should be a design goal? Individualism gets prioritized in the west in particular. But I think a lot of our misery is due to lack of community, the collective. To preserve individual liberty, it will be key to have alternatives and the ability to exit.


Agree completely that we’re seeing values regarding the individual versus the collective being embedded in crypto-systems. I think whether we acknowledge it or not these systems are a-geographic societies. I think the social science side of what’s happening as a result of the intersection between social unrest and tech advancement is radically understudied.

I’m working with a mix of tech and social scientists at the Vienna University of Economics to flesh out an interdisciplinary research roadmap that looks at the two together. There is so much we can learn from the past in terms of social, political and technical change, but there is also a lot of novelty. Complex dynamics we cannot hope to predict but we can at least be more aware as they unfold.

In terms of DAOs, here’s a link to a talk I gave the day after the Cybernetics talk:

This one was focused on diversity at the community level. I think it scratches at some of the topics @s_ben dig into above.

1 Like

Great talk. SourceCred mention starts here in case anyone is curious. Interested to see how SoureCred plays with cadCAD.

So I’m reading Nicholas Taleb’s Antifragile right now, so was primed to be suspect of central planning of economies. When I read this, winced as I imagined, if not 30 million Chinese people starving to death, creeping depression as I toiled in a DAO, followed by unexpected financial ruin.

Was glad to see this on the first slide:

Though still suspicious :face_with_monocle:

Fears mostly abated by the talk. Actually everything seems pretty in line with the concepts of antifragility generally. You’re not relying on models predicting behavior. The emphasis on a ecosystem of DAOs is in line with political decentralization–what Taleb generally argues for, acknowledging that local systems will have their own laws and norms. Using the process of trial and error to find what works, referencing (and borrowing design ideas from) nature/evolution, etc… A part of me is still skeptical about how accurately agents (humans, orgs, etc.) can be modeled. But I think the process of running tons of trial and error simulations, which mirrors biological evolution, and relying on the bare minimum of assumptions and theoretical fragility, should get “in the neighborhood” (as SourceCred already does with reputation), and hopefully at least discover the main failure modes. Seeing failure modes left and right in major crypto projects. Or, if not failure, projects falling into undesirable modes most are unhappy about but can’t change due to on-chain governance systems unable to update the rules (e.g. EOS, Dash).

The concept of energy conservation is making sense. I can see/feel that dynamic working in the Decred DAO, for instance. Changing the consensus rules, is mt Everest, passing a proposal a marathon, etc. There is also a trust barrier often, where you have to expend energy to talk to them without worrying about political consequences.

The question you got about DAOs changing the nature of social interactions. Definitely feel that. Interacting differently than any previous job, though not sure how much to attribute to it being all online vs. the DAO dynamics.

I agree models like this are also just useful as a framework to better understand DAOs generally. I recently read Mario Laul’s article Resource Distribution and Power Dynamics in Decentralized Networks and was fairly mind blown. Not because it had predictive power. But just because I had never encountered sociology before. It was a bit like a light bulb going off, because it allowed me to imagine the entities in the DAO at play in a new way. Like taking a step way back and seeing a big picture I’d not seen before.

Re: communities enforcing norms, just saw this on Twitter, which I thought I you’d appreciate. Community members identifying as cells in an organism.

In the small game of a DAO you can have a bonding curve that exists in isolation and performs an automated market making function. Aragon is building an app for this that allows any DAO to instantly create a market for it’s token. That works fine, but it’s an isolated system.

In the larger game of the crypto markets (like Uniswap and stuff) arbitrage traders are required for the price of Uniswap to achieve parity with other markets. This totally works because the incentives are there, but someone has to build and maintain the plumbing between markets

So I’ve got some experience building and running bots. Had some running on Binance for a while. My trade execution and devops weren’t great (main reason I stopped), but there are now solutions on the market that take care of most of that for you. E.g. have been following hummingbot for a while. They make it pretty easy to deploy arbitrage and market making bots,

1 Like