Exchanging ideas online is the single most important problem we can focus on in mankind. But Internet debate is broken. What would a better online exchange of ideas look like?
The example debates here are all a disagreement on (future) reality. Many of the debates in our world are about policy. In such debates, difference in values also lead to different conclusion. I wonder how to extend the proposed system to such debates
This is such a fantastic post. Did you ever end up writing a follow up? I spent a lot of time over the Christmas break catching up on posts I'd missed over the years but don't remember seeing that one (was mostly focused on the geography ones).
Not a dedicated one. I’ve touched on it here and there, but not focused any article. But I keep thinking about it very actively, and have been doing some things lately on it. I’ll share when ready
I agree that the direction to go is to define a graph with nodes and edges which represents an opinion (or multiple opinions) on the debated issue. I propose to have several types of nodes, for example a node type that represents a value (what we are optimizing for), a node type that represents a proposed policy, and a node type that represents a statement on reality. We can then add edges which connect the different types, e.g. an edge connecting a policy and the value it is supposed to benefit, or an edge that connects a policy to the statement about reality which supports it. Would love to discuss this further.
Perhaps raw materials are more widely produced but their share exceeds most other countries and they seem to have a disproportionate share in their African contracts
Since the article is about a data-driven model, attention to accuracy of its input might be what makes or breaks it.
Tomas focused on 10% error in US population figure but IMO there is one more which is more significant: in 2020 China already had over 1.4 billion inhabitants.
there is no resolvable answer to the question about whether or not the USA and China would go to war over Taiwan - I doubt that even the USA and China know whether they would - I'm reminded of the quote attributed to David Hume that "reason is the slave of the passions" - if push came to shove would reason or passion prevail - the outcome would be very situational and impossible to predict ahead of time - probably the best we can do for issues like this that deal with trying to predict future actions is to surface our assumptions - the key difference between the assumptions of David and Niall seem to be around the future ambitions of China and the USA neither of which it is really possible to 'know' - so, the question makes for an interesting but unresolvable discussion
I hear you and I think you're theoretically right, but wrong in the specifics of the last question.
Eg, I didn't know the US had committed on a paper to actually defend Taiwan. That sounds like a fundamentally interesting fact that supports one of the positions.
Eg, if trade is so important, it deserves more investment in understanding it. Maybe ppl could figure out a historic model that figures out the inverse correlation between trade and war.
American policy on Taiwan/RoC is based on 'strategic ambiguity'. It is designed to dissuade Taiwan/RoC from unilaterally declaring independence AND dissuading China from militarily forcing the reunification of Taiwan with mainland China.
The presumed defence of Taiwan/RoC by America is based on the Taiwan Relations Act, which states that "the United States will make available to Taiwan such defense articles and defense services in such quantity as may be necessary to enable Taiwan to maintain a sufficient self-defense capabilities". However, the decision about the nature and quantity of defense services that America will provide to Taiwan will solely be determined by the President and Congress.
China has also been clear on its future ambitions. It's foreign policy is based on non-interference in the internal affairs of other countries. It does NOT hold itself has a model for other countries to follow and does NOT desire to export its ideology to any other country.
It is a beneficiary and proponent of a global rules-based order. However, the Washington Consensus--the current rules-based order--was designed by and for the benefit of the USA. As the 2nd largest and soon to be largest economy of the world, China now wants a seat at the table. If the West doesn't invite China to the decision making process, then it will simply create its own parallel international institutions such as the Asian Infrastructure Investment Bank.
The Belt and Road Initiative (BRI) is connecting the world to China. 138 countries are participants in BRI. 25 countries count China as their largest export destination and 51 countries count China as their largest import source.
Unfortunately what people say and what they do are two very different things. For a country, you also have the issue that leadership can and does change
What has China been doing??? Has China been trying to export its ideology to other countries? Has China been closing itself to the world? Has China been fighting endless wars?
The reality is that China hasn't fought a single war in over 40 years--since the 1979 Sino-Vietnamese War. And China is opening itself to the world, through the Belt and Road Initiative AND via free trade agreements. In 2021 China joined the Regional Comprehensive Economic Partnership (RCEP)--the largest free trade area in the world, representing 30% of the world's population and 30% of global GDP.
And China has applied to join the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), from which Trump pulled out. The trade agreement that Obama conceived to isolate China will end up isolating America.
Although China's leadership changes, it's policy has been VERY consistent--grow the economy and improve the lives of their citizens. They have lifted 800 million Chinese out of poverty and want to life the remaining 600 million people by 2035.
To be clear, my point was a general one with countless examples throughout history of people saying one thing then doing something different. China may or may not act according to their stated public policies and may or may not act in the future as they have in the recent past, I don't know.
A couple of specific points: it seems evident that Han Chinese have been exporting their ideology to Tibet and Western China. It also seems evident that China has been closing itself off to outside influence from the internet.
Are China and Tibet and Xinjiang independent countries? Or part of China? How can you 'export' something within your own borders?
The Great Firewall of China serves two purposes: 1) censorship from foreign media AND 2) fostering domestic alternatives to Facebook, Google, Twitter, etc. This has allowed China to develop Tencent Video (YouTube), Weibo (Twitter), Qzone (Facebook), WeChat (WhatsApp), Alibaba (Amazon), Baidu (Google), TikTok, etc. At some point I expect to see some of these companies expand to the West--TikTok is already here.
In 2021, China eliminated the requirement for foreign companies to setup joint ventures with domestic companies. In 2019, Chinese tourists made 155 million international return trips. In 2018, the number of international students studying in China grew to 492k and the number of Chinese students studying aboard grew to 662k--hardly evidence of a country closing itself to the outside world.
China has always been VERY clear on the question of Taiwan/RoC. China's priority is the PEACEFUL reunification of Taiwan with the mainland. It doesn't matter how long it takes, China is willing to wait--just like it did with the return of Hong Kong and Macau. China has no desire to invade Taiwan/RoC... UNLESS Taiwan declares independence.
Make no mistake, if Taiwan declares independence China will go to war regardless of the consequences, diplomatic, economic, nuclear, or otherwise.
As usual, Tomas has searched out a better way for debate just as he does with everything that he focuses his mind, and heart, upon. I love visual methods of presenting information, especially inserts of pertinent photos, charts, and graphics that Tomas uses deftly. His summary color chart with the highlights centered in Yellow is a great foundation to work from.
I'm early in the journey, so these things are not ironed out. But here's where I stand today on these:
1. Anybody can write the boxes
2. Pts depend on what you talk about. There are assertions and facts. Facts are independently verified. Assertions are voted.
3. You keep the crazies out through a mechanism of track record. The more you agree with assertions that turn out to be right, and the more you write assertions that ppl agree with or turn out to be right, the more pts you get. The pts you get weight your later votes and the eminence of your new assertions. Crazies don't get exposure because their track record is too bad
This mapping of arguments is a great idea. But for this to work drawing-of-boxes should be automated (text-to-speech, text-analysis, classification, auto-graphics etc.) such that the "argumentation map" is born as discussion happens. I would imagine Presidential Election USA would be very interested in such product. If not for viewers, backstage journalists... Plus AI can bring what it knows before and even participate if asked for opinion (bring a super argument).
I think it is VERY important to pursue the creation of this kind of structured argumentation. Let's imagine that we have perfected it. We should not be discouraged by the fact that a small % of the people on the internet will use it. Most people have a position that is governed by tribal affiliation, and argue to support a tribal position and to inflame their self righteousness. But as always a small number of people usually call the shots and aspire to be rational and improving their batting average is a good thing.
There are several sources of uncertainty that can swing a policy decision: The probability of an outcome of alternative policy options, the relative value of outcomes and then the importance of ideological constraints. Reasonable people can disagree on policy because of disagreements on any or all of these. But if it is made explicit at least one understands that there is no one conclusion that all reasonable people would subscribe to, and this can free up the willingness to muddle through to a good enough solution. The current debates about pandemic policy involves all three of these issues.
Very important subject Tomas! I have seen a bit about Taiwan and its Digital Minister, Audrey Tang, and how it is coming up ways to make the internet less a source for misinformation and polarisation.
I couldn't follow it in detail but it seems they have a Facebook equivalent where the algorithm upgrades when people agree rather than disagree, as well as quite a few more smart things. Sorry I don't have more details but I reckon your research skills would be better than mine. Listening to her talk, I was thinking Taiwan was showing the world how to use technology to make democracy better rather than worse. Tristan Harris from the Centre for Humane Technology and Daniel Schmachtenberger from The Consilience Project are fans. It would be interesting to see where you disagree and agree with them.
I think most debates I encounter these days don't fall into this category - of a debate between two reasonable people, looking for evidence and balanced on what is essentially a judgement call. (e.g. desire for Taiwan vs economic cost).
Instead most debates these days seem to be with people who are down the rabbit-holes, so what you produce as evidence will be dismissed because is from the "MSM", or ignored in favor of some youTube pundit, or the latest Joe Rogan misinformation. :-)
Yes, I used this example as a perfect case of complex argumentation. I want to believe that the case you mention is easier to solve, but I'll take your insight that maybe it's not true. I'll think about it. Thx!
I wish it were easier to solve, unfortunately I think the opposite is true, as its very hard to argue rationally with someone who didn't form their opinion rationally in the first place. We see this with the spreading of missinformation (and dissinformation) in the conspiracy / Q / covid spaces particularly, and the real dangers (and deaths) caused by it.
Thanks for sharing! I really do like this visualization. The first issue my mind went to was the issue of evidence. How do you weigh, value, and visually represent stronger evidence in support of various points in the argument? For example, if a majority of peer-reviewed research supports point X, does that get added weight, visually, vs. "looser" or anecdotal points? And does stronger or more recent research have more value than research with less rigor or recency?
This is at the crux of this! I don't have an answer. But I'm open to ideas.
My guess is that you have ground truth that can be crypto-verified. Eg, what temperature was it in place X? You take the average of 6 thermometers that are crypto-defended and you can know with extremely high confidence what the temperature was.
Assertions combine facts to reach a conclusion.
Today, the validity of assertions is basically the cumulative reputation of the endorsers. Eg if you ask whether climate change is real, the answer is "99.9% of the scientists that follow climate agree that it is real". That is an accumulation of the reputation of the scientists.
So maybe for assertion validity you do that. Ppl's reputation depends on their track record of being right, and the more ppl agree with one assertion, the more it is validated.
The crux there is that you would win a ton of reputation if you turn out to be right AND contrarian.
I like your ideas on this. But there are times when the majority is wrong and the minority is correct. Alfred Wegener (continental drift) and Galileo come to mind.
Perhaps I'm overthinking, but how would you deal with this?
I think the key is to synthesize info and identify surprising agreement and those that are coming to it (ie "experts" of the moment and topic and significance).
1. I would definitely start with Marshall McLuhan. The tool shapes us. "The medium is the message". Your Twitter example unbundles how Twitter UI and Algos fail when it's pushed to an extreme. I think that's the 4th Law of all communication mediums according to McLuhan.
2. However another reason current platforms fail is that the platforms are still incentivized for maximizing engagement not debate SYNTHESIS. They want as many people and bots as possible expressing and competing with shock value. I believe China and Web3.0 will push the "algos" toward "unsubscribe"/"adblock"/"show my tags"... I know - I said China! boo :P
3. In some ways, "synthesis" is more of a ranked list than a "democrazy"... But of course the common denominator person is told that "democratizing" everything is the thing to do and democracy is the best system we've tried, but that's hardly true! Here's are different ways of thinking about that:
A. There are many other types of voting that are better for different purposes.
B. There is Bayesian Truth Serum that is so meta it identifies Truth and Experts. I think this is mindboggling, but still ignored. Sure there are some situations it doesn't apply to, but many where it does. It would be ideal to rank folks into progressive expertise hierarchies or at least get rid of idiots. I'd paste a link but there are many papers to peruse. IMO, this is really the next "AI" - "HGI" human group intelligence. Just turn it into an iterative game... question/answer.
C. For some group decision this paper is relevant. See "Figure 5: Five most accurate and fastest network structures".
wrt comms structure paper. I think they key is that it's for a very specific model of decision making where doctors are convincing each other about the unknown efficacy of a drug. What I get from Figure 5 in that paper is that THE most accurate network had ONLY 2 nodes talking directly to each other and each of those 2 nodes was "informed" by only 2 other nodes. I imagine it as 2 slightly different schools of thought electing highest experts to discuss things in group of 3, then electing THE representative to go and "duke" it out with the adjacent/opposing school(s) of thought. Don't get me wrong, I have no subject matter expertise in this. :) But I think if you stare at that Figure 5 for a bit and imagine how/why the Top 5 Accuracy networks are in their SPECIFIC positions - 1, 2, 3, 4, 5. Then you can start seeing what the key to deriving accuracy may be.
With respect to BTS. I think it's THE mechanism. I feel like Quora and Reddit are already so much better than simple majority that society (imprisoned by social media platform UI "tools) and founders simply stopped iterating. Yet, BTS would 10x expert-identification, idiot elimination, and deriving/synthesizing subjective truths. If I could fast-forward myself to a time when every simple majority poll was supplanted with an extra question, I would. And the question is: "predict what % of poll takers would elect this option". It would be a little bit like Quora but where experts take assessments and create them ITERATIVELY as a very fundamental matter of "discovery" and progress... I sincerely thought Quora was actually going to implement BTS, when they first jumped away from FB's "what are you doing now" preoccupation. Yet, once the masses of later adopters hit ANY communication platform/UI it degrades to somewhere between 0 and harm.
Intuitively "surprising agreement" is tied to Shannon's Information Theory's role of surprise in what constitutions "information" (as opposed to noise or what we now refer to as "data" :) ). I think that's why BTS is intuitively the leap forward for humanity (as opposed to ML which is like a ferrari for reckless drivers). Drazen Prelec would be the guy to track down and talk to.
You may need to use Wayback time machine to get this one...
Ive been thinking about this a bit lately too, in my case I wanted to determine personal decisions around COVID responses (vaccination for children in particular) and it was actually by reading your article that I thought I need to break this down for myself b/c you are persuasive, but you are weighing certain data points more strongly than others (or not at all) and If I do not agree with those then my conclusions may differ. So in a way I was imagining this tool you’re describing to crucially analyse your work! (or more to the point my response to it).
One thing I think that is missing so far that I was focussing on myself was how the supporting evidence for each argument/assertion is factored in, say for e.g, a news article link. This could be weighted (and voted on by the participants by a number of criteria, for e.g. how strongly it supports the assertion, how credible the publication is, how credible the author is (which is probably an amalgam of many factors, but for simplicity just a single metric score would do). So these 3 (+ more I imagine) scores tally up to provide a Strength of Argument Metric (with lots of tweak-ability like weighting some of them, at a global level perhaps if they are considered more impactful - publication may not matter so much, but credibility of Author does etc).
The higher the metric (with more higher calibre evidence), the more strongly that assertion is supported (and then you might even start to visually show the balance tipping for or against when compared to the opposite assertion!) Perhaps though to limit the ability to game the system with a lot of low quality supporting evidence, only the top 10 evidence items are used? These different evidence scores could then be tallied up for each contention (and multiplied by the importance factor of the contention itself) to show ultimately which of the arguments/decisions are the stronger.
I have also been thinking along these lines. Using abductive reasoning, particularly the form called Analysis of Competing Hypotheses (ACH). In addition to weighting the evidence it is necessary to look at confirming, disconfirming, and contradictory evidence. Each piece weighted accordingly. Kepner Tregoe teach a systematic problem solving method that follows this type of reasoning.
One challenge I see in your comment is the urge to quantify things, almost at all cost may I add. I call this very frequent push quantophrenia to describe the try-too-hard focus to put numbers on aspects. One issue is that it usually lays on an underlying belief that a number is better, more precise, more scientific than no number. And this is a belief as in many cases, it is largely incorrect.
In complex cases, such as here the discussion US vs. China used as an example, most things can hardly be quantified. And to do so would be ... to express a personal believe in something, rather than thinking about it and weighing arguments. And this does tend to undermine the strength of the author's proposition.
It does not mean that listing supporting facts or arguments for each point made might not lead to a weighing differing in different views, as you suggest for childrens' Covid-19 vaccination. This would actually be a rather good approach.
But do beware of quantophrenia, the obsession with numbers :-)
The question of a workable format that would work across a wide array of issues is daunting. My first step would be an exhaustive search for ALL the relevant issues at play. What is historically relevant; economic and geographic limits and some concensus on what each party is most likely to value going forward from that particular point in time. Tomas, nowhere in your example is there any discussion of an underlying tension about the dollar vs the yuan. I only point out this as an example of potentially missing one of the relevant issues. So, I would say step one is developing a method of getting all the issues on the table. That's my 2¢.
Counterintuitively, common wisdom when building tech products is to take one specific use case and nail it. And then go from there, expanding to include more use cases. Eg, you can start with a tool that helps academics debate, and go from there.
But yes, I need to have a broader sense of all these use cases before going all in on this
I like the layout of the debate... Fits with my way if breaking down a problem. Didn't quite understand the points / numbers in the yellow boxes and to what extent they influence the deepness of the yellow?
Yeah not great or self-explanatory yet. Thx for noting that.
High-level: the intensity of yellow is meant to convey how important something is, and the #s were meant to convey how important each side thought the argument was. So if each side says it's 10, then they agree this is true and important .
In the last section, the 10 means "The union of China with Taiwan is extremely desired by China", while the -3 means "The union of China with Taiwan is reasonably disfavored by the US".
Feb 1, 2022·edited Feb 1, 2022Liked by Tomas Pueyo
I think showing a visual summary like you did is the way forward. Great idea.
Side question - your debate with Arnold Kling made me remember : I saw somewhere (probably on this blog) that you asked for the US citizenship.
This really surprises me, for I assumed that, because you wrote these articles about the end of nation-states, you were maybe libertarian leaning - or at least that you plan your future today according to your own predictions (which is that in the future nation-states will fight to the death to keep the status quo).
So there is a big risk that many nation states will try to restrain the freedom and mobility of their most mobile citizen, in one way or another.
And if it is so, why did you choose to take on the only* nationality on Earth that will tax you no matter where you live, prevent you from opening bank accounts in many countries, and burden you with paperwork to fill every year even if you don't have any more ties with the US ?
Don't you want to keep this freedom to choose to let go of a state, if it becomes too overarching ?
*yes Eretria does that too, theoretically, but... well you know what I mean :)
Feb 3, 2022·edited Feb 4, 2022Liked by Tomas Pueyo
There are plenty of EU countries that tax less than the US, especially with their special tax regimes for foreigners, as you noted yourself in your article.
Also, you are not protected against the possibility of becoming wealthy, especially given your skills :) .
If you manage to do it, I believe you will come to regret your decision.
I have quite a few friends who renounced their US citizenship, and it was not a good customer experience.
When I see this (as well as the exit tax you have to pay when you renounce), I can't help but see the parallel with the Middle Age, when you were the property of your lord, and had to ask his permission in order to leave the fiefdom, and usually had to buy him back your freedom.
With a system like this, you are the property of your government. That's the only reason I'm not an American right now, I think this system is revolting and destroy too many essential liberties - the right to vote with your feet being the most important one.
Let's hope I'm wrong, and that you will find more benefits to it than disavantages.
Loved the discussion topology, before I saw your images I started visualizing a cluster diagram with larger and smaller circles representing their relevance, I think that would also work very well.
I like this idea, and would add thar I think larger and smaller sized circles would be easier to understand at a glance than the intensity of a colour (in this case, yellow).
The example debates here are all a disagreement on (future) reality. Many of the debates in our world are about policy. In such debates, difference in values also lead to different conclusion. I wonder how to extend the proposed system to such debates
This is actually easier for things like "what should we do about X topic". I hope to take examples of that in the future
This is such a fantastic post. Did you ever end up writing a follow up? I spent a lot of time over the Christmas break catching up on posts I'd missed over the years but don't remember seeing that one (was mostly focused on the geography ones).
Thank you!
Not a dedicated one. I’ve touched on it here and there, but not focused any article. But I keep thinking about it very actively, and have been doing some things lately on it. I’ll share when ready
I agree that the direction to go is to define a graph with nodes and edges which represents an opinion (or multiple opinions) on the debated issue. I propose to have several types of nodes, for example a node type that represents a value (what we are optimizing for), a node type that represents a proposed policy, and a node type that represents a statement on reality. We can then add edges which connect the different types, e.g. an edge connecting a policy and the value it is supposed to benefit, or an edge that connects a policy to the statement about reality which supports it. Would love to discuss this further.
Several added boxes
China makes most of our stuff including chips
We buy most of our stuff from China including chips
China has billion people
We have 360 million
Chinas middle class is rising as ours is declining
Soon China will just need to sell to the rest of the world and themselves
China also produces and refines most of the worlds raw materials
We need raw materials like iron and rare earth elements
China has a military and operational control of the Taiwan area
Our military leaders have noted under no scenario could our military win in a limited conflict with China
An all out war would not be acceptable to either side
Tianmin Square, Hong Kong and Vietnam provide examples of use and limits of power
You are right in most of these, even if maybe somewhat imprecise in some.
Eg, Taiwan does most of our chips AFAIK
Eg, US pop is 330M
Eg, US middle class is stagnating as a share of income, but is growing in absolute terms
eg, China does not produce and refine most of the world's raw materials. They do for rare earth elements.
Where I disagree more fundamentally is that your take on China is a bit one-sided. They have many weaknesses too. I'll cover them in a future article.
Agree that all have weaknesses
Perhaps raw materials are more widely produced but their share exceeds most other countries and they seem to have a disproportionate share in their African contracts
I await your next article
Since the article is about a data-driven model, attention to accuracy of its input might be what makes or breaks it.
Tomas focused on 10% error in US population figure but IMO there is one more which is more significant: in 2020 China already had over 1.4 billion inhabitants.
there is no resolvable answer to the question about whether or not the USA and China would go to war over Taiwan - I doubt that even the USA and China know whether they would - I'm reminded of the quote attributed to David Hume that "reason is the slave of the passions" - if push came to shove would reason or passion prevail - the outcome would be very situational and impossible to predict ahead of time - probably the best we can do for issues like this that deal with trying to predict future actions is to surface our assumptions - the key difference between the assumptions of David and Niall seem to be around the future ambitions of China and the USA neither of which it is really possible to 'know' - so, the question makes for an interesting but unresolvable discussion
I hear you and I think you're theoretically right, but wrong in the specifics of the last question.
Eg, I didn't know the US had committed on a paper to actually defend Taiwan. That sounds like a fundamentally interesting fact that supports one of the positions.
Eg, if trade is so important, it deserves more investment in understanding it. Maybe ppl could figure out a historic model that figures out the inverse correlation between trade and war.
American policy on Taiwan/RoC is based on 'strategic ambiguity'. It is designed to dissuade Taiwan/RoC from unilaterally declaring independence AND dissuading China from militarily forcing the reunification of Taiwan with mainland China.
The presumed defence of Taiwan/RoC by America is based on the Taiwan Relations Act, which states that "the United States will make available to Taiwan such defense articles and defense services in such quantity as may be necessary to enable Taiwan to maintain a sufficient self-defense capabilities". However, the decision about the nature and quantity of defense services that America will provide to Taiwan will solely be determined by the President and Congress.
China has also been clear on its future ambitions. It's foreign policy is based on non-interference in the internal affairs of other countries. It does NOT hold itself has a model for other countries to follow and does NOT desire to export its ideology to any other country.
It is a beneficiary and proponent of a global rules-based order. However, the Washington Consensus--the current rules-based order--was designed by and for the benefit of the USA. As the 2nd largest and soon to be largest economy of the world, China now wants a seat at the table. If the West doesn't invite China to the decision making process, then it will simply create its own parallel international institutions such as the Asian Infrastructure Investment Bank.
The Belt and Road Initiative (BRI) is connecting the world to China. 138 countries are participants in BRI. 25 countries count China as their largest export destination and 51 countries count China as their largest import source.
Unfortunately what people say and what they do are two very different things. For a country, you also have the issue that leadership can and does change
What has China been doing??? Has China been trying to export its ideology to other countries? Has China been closing itself to the world? Has China been fighting endless wars?
The reality is that China hasn't fought a single war in over 40 years--since the 1979 Sino-Vietnamese War. And China is opening itself to the world, through the Belt and Road Initiative AND via free trade agreements. In 2021 China joined the Regional Comprehensive Economic Partnership (RCEP)--the largest free trade area in the world, representing 30% of the world's population and 30% of global GDP.
And China has applied to join the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), from which Trump pulled out. The trade agreement that Obama conceived to isolate China will end up isolating America.
Although China's leadership changes, it's policy has been VERY consistent--grow the economy and improve the lives of their citizens. They have lifted 800 million Chinese out of poverty and want to life the remaining 600 million people by 2035.
To be clear, my point was a general one with countless examples throughout history of people saying one thing then doing something different. China may or may not act according to their stated public policies and may or may not act in the future as they have in the recent past, I don't know.
A couple of specific points: it seems evident that Han Chinese have been exporting their ideology to Tibet and Western China. It also seems evident that China has been closing itself off to outside influence from the internet.
Are China and Tibet and Xinjiang independent countries? Or part of China? How can you 'export' something within your own borders?
The Great Firewall of China serves two purposes: 1) censorship from foreign media AND 2) fostering domestic alternatives to Facebook, Google, Twitter, etc. This has allowed China to develop Tencent Video (YouTube), Weibo (Twitter), Qzone (Facebook), WeChat (WhatsApp), Alibaba (Amazon), Baidu (Google), TikTok, etc. At some point I expect to see some of these companies expand to the West--TikTok is already here.
In 2021, China eliminated the requirement for foreign companies to setup joint ventures with domestic companies. In 2019, Chinese tourists made 155 million international return trips. In 2018, the number of international students studying in China grew to 492k and the number of Chinese students studying aboard grew to 662k--hardly evidence of a country closing itself to the outside world.
I am not quite sure what the point you are making is. Are you able to clarify?
China has always been VERY clear on the question of Taiwan/RoC. China's priority is the PEACEFUL reunification of Taiwan with the mainland. It doesn't matter how long it takes, China is willing to wait--just like it did with the return of Hong Kong and Macau. China has no desire to invade Taiwan/RoC... UNLESS Taiwan declares independence.
Make no mistake, if Taiwan declares independence China will go to war regardless of the consequences, diplomatic, economic, nuclear, or otherwise.
Hey John, this is a place where we debate thoughtfully and respectfully.
This is clearly a topic you're passionate about, but let's keep in mind total respect for others and humility. We never know when we might be wrong.
As usual, Tomas has searched out a better way for debate just as he does with everything that he focuses his mind, and heart, upon. I love visual methods of presenting information, especially inserts of pertinent photos, charts, and graphics that Tomas uses deftly. His summary color chart with the highlights centered in Yellow is a great foundation to work from.
Thank you Madjag, that is kind!
Nice idea but who gets to write the boxes and who assigns points?? And how to keep the crazies out??
I'm early in the journey, so these things are not ironed out. But here's where I stand today on these:
1. Anybody can write the boxes
2. Pts depend on what you talk about. There are assertions and facts. Facts are independently verified. Assertions are voted.
3. You keep the crazies out through a mechanism of track record. The more you agree with assertions that turn out to be right, and the more you write assertions that ppl agree with or turn out to be right, the more pts you get. The pts you get weight your later votes and the eminence of your new assertions. Crazies don't get exposure because their track record is too bad
This mapping of arguments is a great idea. But for this to work drawing-of-boxes should be automated (text-to-speech, text-analysis, classification, auto-graphics etc.) such that the "argumentation map" is born as discussion happens. I would imagine Presidential Election USA would be very interested in such product. If not for viewers, backstage journalists... Plus AI can bring what it knows before and even participate if asked for opinion (bring a super argument).
I think it is VERY important to pursue the creation of this kind of structured argumentation. Let's imagine that we have perfected it. We should not be discouraged by the fact that a small % of the people on the internet will use it. Most people have a position that is governed by tribal affiliation, and argue to support a tribal position and to inflame their self righteousness. But as always a small number of people usually call the shots and aspire to be rational and improving their batting average is a good thing.
There are several sources of uncertainty that can swing a policy decision: The probability of an outcome of alternative policy options, the relative value of outcomes and then the importance of ideological constraints. Reasonable people can disagree on policy because of disagreements on any or all of these. But if it is made explicit at least one understands that there is no one conclusion that all reasonable people would subscribe to, and this can free up the willingness to muddle through to a good enough solution. The current debates about pandemic policy involves all three of these issues.
Raymond Neutra MD
Very important subject Tomas! I have seen a bit about Taiwan and its Digital Minister, Audrey Tang, and how it is coming up ways to make the internet less a source for misinformation and polarisation.
I couldn't follow it in detail but it seems they have a Facebook equivalent where the algorithm upgrades when people agree rather than disagree, as well as quite a few more smart things. Sorry I don't have more details but I reckon your research skills would be better than mine. Listening to her talk, I was thinking Taiwan was showing the world how to use technology to make democracy better rather than worse. Tristan Harris from the Centre for Humane Technology and Daniel Schmachtenberger from The Consilience Project are fans. It would be interesting to see where you disagree and agree with them.
cheers
Tony
Found these
https://consilienceproject.org/taiwans-digital-democracy/#textref-22
https://theconversation.com/hacking-the-pandemic-how-taiwans-digital-democracy-holds-covid-19-at-bay-145023
I think most debates I encounter these days don't fall into this category - of a debate between two reasonable people, looking for evidence and balanced on what is essentially a judgement call. (e.g. desire for Taiwan vs economic cost).
Instead most debates these days seem to be with people who are down the rabbit-holes, so what you produce as evidence will be dismissed because is from the "MSM", or ignored in favor of some youTube pundit, or the latest Joe Rogan misinformation. :-)
Yes, I used this example as a perfect case of complex argumentation. I want to believe that the case you mention is easier to solve, but I'll take your insight that maybe it's not true. I'll think about it. Thx!
I wish it were easier to solve, unfortunately I think the opposite is true, as its very hard to argue rationally with someone who didn't form their opinion rationally in the first place. We see this with the spreading of missinformation (and dissinformation) in the conspiracy / Q / covid spaces particularly, and the real dangers (and deaths) caused by it.
Maybe the system should not accept arguments too small. 2ct minimum
Certainly worth a try …..as you say…the current system is rather useless
Thanks for sharing! I really do like this visualization. The first issue my mind went to was the issue of evidence. How do you weigh, value, and visually represent stronger evidence in support of various points in the argument? For example, if a majority of peer-reviewed research supports point X, does that get added weight, visually, vs. "looser" or anecdotal points? And does stronger or more recent research have more value than research with less rigor or recency?
This is at the crux of this! I don't have an answer. But I'm open to ideas.
My guess is that you have ground truth that can be crypto-verified. Eg, what temperature was it in place X? You take the average of 6 thermometers that are crypto-defended and you can know with extremely high confidence what the temperature was.
Assertions combine facts to reach a conclusion.
Today, the validity of assertions is basically the cumulative reputation of the endorsers. Eg if you ask whether climate change is real, the answer is "99.9% of the scientists that follow climate agree that it is real". That is an accumulation of the reputation of the scientists.
So maybe for assertion validity you do that. Ppl's reputation depends on their track record of being right, and the more ppl agree with one assertion, the more it is validated.
The crux there is that you would win a ton of reputation if you turn out to be right AND contrarian.
I like your ideas on this. But there are times when the majority is wrong and the minority is correct. Alfred Wegener (continental drift) and Galileo come to mind.
Perhaps I'm overthinking, but how would you deal with this?
If you insert betting markets, those right and contrarian could become rich.
I think the key is to synthesize info and identify surprising agreement and those that are coming to it (ie "experts" of the moment and topic and significance).
1. I would definitely start with Marshall McLuhan. The tool shapes us. "The medium is the message". Your Twitter example unbundles how Twitter UI and Algos fail when it's pushed to an extreme. I think that's the 4th Law of all communication mediums according to McLuhan.
2. However another reason current platforms fail is that the platforms are still incentivized for maximizing engagement not debate SYNTHESIS. They want as many people and bots as possible expressing and competing with shock value. I believe China and Web3.0 will push the "algos" toward "unsubscribe"/"adblock"/"show my tags"... I know - I said China! boo :P
3. In some ways, "synthesis" is more of a ranked list than a "democrazy"... But of course the common denominator person is told that "democratizing" everything is the thing to do and democracy is the best system we've tried, but that's hardly true! Here's are different ways of thinking about that:
A. There are many other types of voting that are better for different purposes.
https://en.wikipedia.org/wiki/Approval_voting#Compliance_with_voting_system_criteria
B. There is Bayesian Truth Serum that is so meta it identifies Truth and Experts. I think this is mindboggling, but still ignored. Sure there are some situations it doesn't apply to, but many where it does. It would be ideal to rank folks into progressive expertise hierarchies or at least get rid of idiots. I'd paste a link but there are many papers to peruse. IMO, this is really the next "AI" - "HGI" human group intelligence. Just turn it into an iterative game... question/answer.
C. For some group decision this paper is relevant. See "Figure 5: Five most accurate and fastest network structures".
http://www.kevinzollman.com/uploads/5/0/3/6/50361245/zollman_-_communication_structure.pdf
Very interesting. Looked into the BTS. There's something interesting there.
Also the paper on comms structure, from a quick skim, seems more disheartening than anything. We know so little about this...
wrt comms structure paper. I think they key is that it's for a very specific model of decision making where doctors are convincing each other about the unknown efficacy of a drug. What I get from Figure 5 in that paper is that THE most accurate network had ONLY 2 nodes talking directly to each other and each of those 2 nodes was "informed" by only 2 other nodes. I imagine it as 2 slightly different schools of thought electing highest experts to discuss things in group of 3, then electing THE representative to go and "duke" it out with the adjacent/opposing school(s) of thought. Don't get me wrong, I have no subject matter expertise in this. :) But I think if you stare at that Figure 5 for a bit and imagine how/why the Top 5 Accuracy networks are in their SPECIFIC positions - 1, 2, 3, 4, 5. Then you can start seeing what the key to deriving accuracy may be.
With respect to BTS. I think it's THE mechanism. I feel like Quora and Reddit are already so much better than simple majority that society (imprisoned by social media platform UI "tools) and founders simply stopped iterating. Yet, BTS would 10x expert-identification, idiot elimination, and deriving/synthesizing subjective truths. If I could fast-forward myself to a time when every simple majority poll was supplanted with an extra question, I would. And the question is: "predict what % of poll takers would elect this option". It would be a little bit like Quora but where experts take assessments and create them ITERATIVELY as a very fundamental matter of "discovery" and progress... I sincerely thought Quora was actually going to implement BTS, when they first jumped away from FB's "what are you doing now" preoccupation. Yet, once the masses of later adopters hit ANY communication platform/UI it degrades to somewhere between 0 and harm.
Intuitively "surprising agreement" is tied to Shannon's Information Theory's role of surprise in what constitutions "information" (as opposed to noise or what we now refer to as "data" :) ). I think that's why BTS is intuitively the leap forward for humanity (as opposed to ML which is like a ferrari for reckless drivers). Drazen Prelec would be the guy to track down and talk to.
You may need to use Wayback time machine to get this one...
(2008) Rebecca Weiss, Drazen Prelec. "An interactive tool to optimally aggregate crowd wisdom".(http://www.princeton.edu/~decconf/FinalPapers/WeissPaper.pdf)
Ive been thinking about this a bit lately too, in my case I wanted to determine personal decisions around COVID responses (vaccination for children in particular) and it was actually by reading your article that I thought I need to break this down for myself b/c you are persuasive, but you are weighing certain data points more strongly than others (or not at all) and If I do not agree with those then my conclusions may differ. So in a way I was imagining this tool you’re describing to crucially analyse your work! (or more to the point my response to it).
One thing I think that is missing so far that I was focussing on myself was how the supporting evidence for each argument/assertion is factored in, say for e.g, a news article link. This could be weighted (and voted on by the participants by a number of criteria, for e.g. how strongly it supports the assertion, how credible the publication is, how credible the author is (which is probably an amalgam of many factors, but for simplicity just a single metric score would do). So these 3 (+ more I imagine) scores tally up to provide a Strength of Argument Metric (with lots of tweak-ability like weighting some of them, at a global level perhaps if they are considered more impactful - publication may not matter so much, but credibility of Author does etc).
The higher the metric (with more higher calibre evidence), the more strongly that assertion is supported (and then you might even start to visually show the balance tipping for or against when compared to the opposite assertion!) Perhaps though to limit the ability to game the system with a lot of low quality supporting evidence, only the top 10 evidence items are used? These different evidence scores could then be tallied up for each contention (and multiplied by the importance factor of the contention itself) to show ultimately which of the arguments/decisions are the stronger.
This makes sense. Yes, I think a solution would go in this direction!
And I'm so glad my articles were triggers for you to check my points!
Sorry that should have read CRITICALLY analyse your work obviously! Can't wait to get my hands on this tool! When can we play with the prototype? ;)
Ugh, I can’t focus on this now…
I have also been thinking along these lines. Using abductive reasoning, particularly the form called Analysis of Competing Hypotheses (ACH). In addition to weighting the evidence it is necessary to look at confirming, disconfirming, and contradictory evidence. Each piece weighted accordingly. Kepner Tregoe teach a systematic problem solving method that follows this type of reasoning.
Good comment on an excellent proposition.
One challenge I see in your comment is the urge to quantify things, almost at all cost may I add. I call this very frequent push quantophrenia to describe the try-too-hard focus to put numbers on aspects. One issue is that it usually lays on an underlying belief that a number is better, more precise, more scientific than no number. And this is a belief as in many cases, it is largely incorrect.
In complex cases, such as here the discussion US vs. China used as an example, most things can hardly be quantified. And to do so would be ... to express a personal believe in something, rather than thinking about it and weighing arguments. And this does tend to undermine the strength of the author's proposition.
It does not mean that listing supporting facts or arguments for each point made might not lead to a weighing differing in different views, as you suggest for childrens' Covid-19 vaccination. This would actually be a rather good approach.
But do beware of quantophrenia, the obsession with numbers :-)
The question of a workable format that would work across a wide array of issues is daunting. My first step would be an exhaustive search for ALL the relevant issues at play. What is historically relevant; economic and geographic limits and some concensus on what each party is most likely to value going forward from that particular point in time. Tomas, nowhere in your example is there any discussion of an underlying tension about the dollar vs the yuan. I only point out this as an example of potentially missing one of the relevant issues. So, I would say step one is developing a method of getting all the issues on the table. That's my 2¢.
Thx Weston!
Counterintuitively, common wisdom when building tech products is to take one specific use case and nail it. And then go from there, expanding to include more use cases. Eg, you can start with a tool that helps academics debate, and go from there.
But yes, I need to have a broader sense of all these use cases before going all in on this
I like the layout of the debate... Fits with my way if breaking down a problem. Didn't quite understand the points / numbers in the yellow boxes and to what extent they influence the deepness of the yellow?
Yeah not great or self-explanatory yet. Thx for noting that.
High-level: the intensity of yellow is meant to convey how important something is, and the #s were meant to convey how important each side thought the argument was. So if each side says it's 10, then they agree this is true and important .
In the last section, the 10 means "The union of China with Taiwan is extremely desired by China", while the -3 means "The union of China with Taiwan is reasonably disfavored by the US".
I think showing a visual summary like you did is the way forward. Great idea.
Side question - your debate with Arnold Kling made me remember : I saw somewhere (probably on this blog) that you asked for the US citizenship.
This really surprises me, for I assumed that, because you wrote these articles about the end of nation-states, you were maybe libertarian leaning - or at least that you plan your future today according to your own predictions (which is that in the future nation-states will fight to the death to keep the status quo).
So there is a big risk that many nation states will try to restrain the freedom and mobility of their most mobile citizen, in one way or another.
And if it is so, why did you choose to take on the only* nationality on Earth that will tax you no matter where you live, prevent you from opening bank accounts in many countries, and burden you with paperwork to fill every year even if you don't have any more ties with the US ?
Don't you want to keep this freedom to choose to let go of a state, if it becomes too overarching ?
*yes Eretria does that too, theoretically, but... well you know what I mean :)
Very good question. I struggled about that question for a very long time.
Eventually, I reached the conclusion that:
- I will most likely live either in the US or in EU countries
- EU countries always tax more than the US, so I would not end up paying (much) US taxes
- The upside of a US nationality in terms of derisking the next few decades is higher than the cost
I might have been wrong.
There are plenty of EU countries that tax less than the US, especially with their special tax regimes for foreigners, as you noted yourself in your article.
Also, you are not protected against the possibility of becoming wealthy, especially given your skills :) .
If you manage to do it, I believe you will come to regret your decision.
I have quite a few friends who renounced their US citizenship, and it was not a good customer experience.
The US even used COVID as an excuse to stop processing renunciations for 2 years, and the Association of Accidental Americans is suing the federal state for this : www.imidaily.com/north-america/expat-group-sues-us-state-dept-over-suspension-of-renunciation-services-violation-of-5th-amendment/
When I see this (as well as the exit tax you have to pay when you renounce), I can't help but see the parallel with the Middle Age, when you were the property of your lord, and had to ask his permission in order to leave the fiefdom, and usually had to buy him back your freedom.
With a system like this, you are the property of your government. That's the only reason I'm not an American right now, I think this system is revolting and destroy too many essential liberties - the right to vote with your feet being the most important one.
Let's hope I'm wrong, and that you will find more benefits to it than disavantages.
Loved the discussion topology, before I saw your images I started visualizing a cluster diagram with larger and smaller circles representing their relevance, I think that would also work very well.
I like this idea, and would add thar I think larger and smaller sized circles would be easier to understand at a glance than the intensity of a colour (in this case, yellow).