Thread by @sfiscience

Today's Applied Complexity Symposium begins with a panel on Computation and Complex Economies, feat. Rob Axtell, Joshua Epstein, @C4COMPUTATION, Blake LeBaron, John Miller, & @MelMitchell1

(Follow this thread for live coverage all day...)

#complexsystems #economics

4 Principles of Collective #Computation:

1 Ground truth (objective reality)
2 Effective ground truth (what we agree on, whether or not it is accurate)
3 #Information can be collectively encoded in #networks
4 Outputs are a product of collective dynamics

@C4COMPUTATION right now

"I think #economics really needs to adopt a generative framework with agent-based models of cognitively plausible software people...[and] we don't choose to fear."

- SFI'S Joshua Epstein (@GeorgeMasonU) on modeling emotions at today's ACtioN symposium:
Santafe.edu/action

"Not every dynamical system is computing something. There has to be an interpreter."
- @MelMitchell1

"For adaptive #systems, we don't know what #computation is yet. There has to be computation FOR. And I don't think that's thought about enough in CS."
- @C4COMPUTATION

CORRECTION: Epstein is at @nyuniversity not @GeorgeMasonU. Those responsible have been sacked.

What is the #economy?

"The 20th Century definition is that it is a resource-allocation mechanism. That sounds somewhat archaic today. But as an actor within an economy, it's an NP-hard problem to know how to act."

- SFI's Rob Axtell (@GeorgeMasonU)

"You can say that the #brain is an economy. I think the opposite is also true: that the economy is a #brain. Does the #economy have its own 'cognitive biases' that lead to inefficient resource allocation?"

- @MelMitchell1

Does using AI-generated economic agent-based models answer the questions we think it does?

"I think you can beat humans and augment humans, but you can't explain humans [with neural nets]. This is all totally irrelevant to social science."

- SFI's Joshua Epstein (@nyuniversity)

"Your brain makes computations while you're trail running. But you have no access to those computations. MOST of the models we make in #science are totally incomprehensible, and what we do as scientists is do dimension reduction to extract causal relationships."

- @C4COMPUTATION

"I define #error as the deviation from the maximal theoretical structure that could be abstracted from the environment."

@C4COMPUTATION on what we even mean by "error" in discussing #computation

"Are institutions evolved to correct these [individual] biases? Or do they magnify them?"

SFI's Blake LeBaron (@BrandeisU) on the efficacy of zero-intelligence models in understanding collective behavior

"I think when you ask about [what the economy optimizes for] it is a #policy question. It is much less expensive to experiment in artificial systems...policy simulators, broadly construed."

- SFI's Rob Axtell (@GeorgeMasonU) on whether we have to serve an ocean-plastic algorithm

Next up today: #Physics and Economic #Systems, featuring @cshalizi @RMaria_drc @ole_b_peters & David Wolpert.

Opening shot from Cosma:
"I don't actually think physics has lent that much to our understanding of the #economy. The repertoire is ridiculously too limited."

"#Physics started making progress when we actually started *seeing* the world. In #economics we have one advantage [over quantum physics], and that is, we are the micro-scale."

@RMaria_drc on the importance of observing the #data for good #econophysics

"If you take an economics textbook and do something useful with it, and throw it out a window, it will follow a parabola."

SFI's @ole_b_peters (@LdnMathLab) on the relationship between #economics & #physics

1/ "There are technologies converging...one of them is #blockchain, which enables #frictionless binding contracts. The default mode is that they are publicly visible. As soon as you have [these], in many situations, a third party can come in, offer new #contracts..."

2/ "...and completely screw up those contracts. If I try to #money pump one contract someone else can come in and try to money pump me. This is the space of all possible contracts. And we don't know how to model this. It's completely #unforeseeable with the tools that we have..."

3/ "The second technology is reinforcement learning #algorithms. Let them loose on your behalf trying to find contracts to money pump, just like #AlphaGo. All of this is likely to occur in 5-10 years & if it does it'll make all this talk of the #Singularity seem small-minded..."

4/ "[Therefore, the #future of the #economy is] way beyond #physics."

- SFI's David Wolpert

"We have to stop copy-pasting things from #physics into #economics. We have to stop reinventing the wheel. If we want economists to listen to us [physicists], we have to read them."
- @RMaria_drc

About today's symposium:
santafe.edu/action

"Physicists are arrogant and economists are smug. And I can say that because I've been both. Physicists think their models can solve everything...that was a problem in the early days of SFI. [But] physicists are quite good at taking tools from all over the place."
- David Wolpert

"We don't care in the economy about just sending Information from Point A to Point B. We care about transforming it. So information in the Shannon sense isn't useful to us here [in #econophysics]."
- David Wolpert

"We're actually living in the *aftermath* of the #singularity. Those artificial intelligences are the institutions, the corporations...they've been in charge for a hundred years, and we're living in the #postsingularity nightmare."

- @cshalizi (@CarnegieMellon)

"The #theory tells you want #data to collect. The data doesn't give you the theory. I think this whole thing about 'data-first' is extremely misguided."

- SFI's Joshua Epstein (@nyuniversity)

Our next panel: The Economic Organism
featuring @cmoncap @Scott_E_Page @rajivatbarnard & Geoffrey West

santafe.edu/action #complexsystems

"The question on the table to me about SFI is, does this approach help at a pragmatic level? I want to get back to #coarsegraining. It's not that my model beats your model. It's a both-and thing. If we stick to one model, we're bound to fail."
- @Scott_E_Page

#pluralism

"Complexity #economics is still peripheral to the discipline. The cynical view the pros climbed the wrong mountain, and it's too difficult to climb down again. But my view is that the models are too easy to construct; the referees don't know which ones are good."
@rajivatbarnard

"The most pragmatic problem is the future of this planet, and its survival. And I think one of the great disgraces of academic economists is that they're not at the forefront of this. It's morally reprehensible that they're not leading the way."
- Geoffrey West

"There's a movement now in Michigan against letting people order Starbucks from an app. You don't talk to the person behind the counter. Eric [Beinhocker] made this argument that we're social animals. But if we allow technology to isolate us, that all goes away."
- @Scott_E_Page

"A lot of economic research has been about understanding markets. But a lot of economic activity goes on in networks. Economists have been limited by the availability of data. I think [they] are adjusting, but it's slow."
- @cmoncap

"The important point about physics models is that they are abstractions. If Galileo had been a biologist, instead of concluding that everything falls at the same rate, he'd have done a great taxonomy of the different ways that things fall."
- Geoffrey West
santafe.edu/ACTioN

"Can we anticipate how some systems will be gamed or taken advantage of and collapse? (This is something the medical profession understands very well.) What would it mean to construct a clever intelligent system that explores how policies can be gamed?"

- W. Brian Arthur

"In neoclassical #economics, markets and governments are thought of as independent of each other. I would see this instead as an #ecology. In this view, the market creates competition to solve human problems, and government enables the cooperation to do this."
- @EricBeinhocker

"The market doesn't talk about who gets what, the distribution of profits and rents. And it doesn't talk about externalities. We have to know, if I default, who is it going to hurt?"

@JacksonmMatt on the Economic Architecture panel with W. Brian Arthur @EricBeinhocker @AKStanger

"If we get more voices in the room, we can think more creatively about economics. And that's really what we're trying to do at SFI."

- @AKStanger on modeling a #moraleconomy & replacing Homo economicus with Sapiens integra, which cares about maximizing positive externalities

"Having #wealth is not all that useful if people can't have #access to it. We're moving from problems of #production to #problems of access, from #efficiency to who gets what. How do we get by in an era in which #distribution of wealth is a dominant problem?"

- W. Brian Arthur

"Unless the #politics are willing to make long-term, very costly decisions, nothing is going to happen [with #climate accords]. We have to have a better understanding of the interaction between politics and #economics."
- @JacksonmMatt

"All of the developed countries are dealing with the same challenges from #technology, but some of them have held on to their middle classes, and some haven't. It appears to be a policy issue."
- @EricBeinhocker

Our last panel of the #Complexity #Economics Symposium is "From Theory to Application," featuring @mjmauboussin @honeybeecap Dario Villani @B3_MillerValue, moderated by @PaulJDavies of @WSJ:

"Every model of #abstraction you make, your models get prettier and prettier, but the world you're describing gets farther and farther away."

- @honeybeecap on the delicate balance involved in actually applying unifying frameworks in the market

"The further out you look, the more #uncertainty. So there are more opportunities out there, the longer your time horizons... Broadly speaking, #complexity #economics is about being less model-dependent."
- @B3_MillerValue

"Our ability to #forecast is naturally limited. The difference between me with complicated models and someone else without is #timing."

- Dario Villani

"It's much harder to outperform [the market] short-term. It's easy to outperform long-term."

"We have a big gap between perceived risk and real risk."

- @B3_MillerValue

Santafe.edu/action

"It used to be to ask, what are you against? The question [for investors] now is, what are you FOR? As soon as you ask this, you get a really strong nonpolitical convergence. Nobody doesn't want clean water for their children."

- @honeybeecap