knowing reality as it is
How can we know what reality really is?
There is a website that should exist, which would be complementary to Wikipedia in its goal of collecting and organizing human knowledge, but distinct from it in format. Where Wikipedia is structured as an encyclopedia and is meant to handle facts with a neutral overview account, the proposed site would instead be structured as graphical branching chains of ideas and would be meant to handle argumentative knowledge - those things we believe in on the basis of a chain of arguments for their probable truth: x is so because y is so, and y is so because z is so, etc. This kind of knowledge requires, for real understanding, not only a clear statement of a specific claim itself, but an understanding of the claims it depends on as well as an understanding of the counter-claims it is competing with (and they can't be understood without understanding their underlying claims, etc.) Most of our most important 'knowledge' is argumentative in this sense, including political, religious, and cultural ideas, and even most of scientific knowledge, since science involves not merely empirical facts but the interpretation of those facts and the development of theories about them. We need better formats for making the necessary contextual information easily accessible.
Almost all of our public discussion is argumentative in nature, but nearly all of it takes place without sufficient context, and in particular without any easy way to fully examine the underlying claims of the argument being made, or the counter-arguments that could be made against it. To reasonably assess a claim, one has to do all one's own research across dozens of websites, reading opinion pieces by all sides, research papers where applicable, and tracing all the arguments in every one back through their underlying claims and sources of alleged evidence. Even if you had the time or energy to do so, there is no way to guarantee that your research has turned up everything relevant, or assessed it accurately. And without adequate context, or even the ability to know when you have achieved adequate context, there is no such thing as a genuinely reasonable opinion on any even moderately complex topic - if you do not know what the strongest arguments are against what you think is true, or even all the plausible arguments against what you think is true, you cannot know that what you think is probably true, is probably true. For deep questions, nothing short of the equivalent of a Ph.D program is sufficient to hold an opinion you can justifiably believe is adequately informed and nuanced. There can be no meaningfully informed populace in any world with more than a handful of complex topics, unless there is some structure capable of reducing those complexes of ideas to easily absorbable and assessable forms.
And yet all of the necessary information exists, scattered throughout the world, the internet, and the minds of various individuals. The problem is one of collecting and organizing the necessary information and rendering it reliable, understandable, and accessible. It's a question of format and social mechanisms. Those are problems that can be solved. I will discuss approaches that seem to me to be plausible first-order solutions in following posts.
What I would like to stress right now is the importance of solving this problem - of making reliable, nuanced understandings of complex questions more accessible and more common - and the consequences of not solving it, which I believe are many and possibly dire.
We are social creatures who make collective decisions (whether that is the explicit structure of our political systems, or simply the underlying social reality within which even dictators have to be careful not to overstep the bounds of what their society will accept). Our collective decisions about the range of acceptable beliefs and behaviors are conditioned by our understanding of what reality is and how it works, by how confident we are in our assessment of what's real, and also by how confident we are in our assessment of what the public in general believes is real. If there is significant ambiguity in our individual models of reality, or significant discrepancy between competing models of reality in the general public, and there is no reliable public method for assessing the relative strengths of claims, then a number of problems arise.
One is that unscrupulous actors can take advantage of the ambiguity and confusion to simply assert that certain claims are false, promote their own preferred claims as alternative truths, and, with enough charisma or perception-manipulating genius, get away with reality-distortion on a large scale because there exists no large social consensus (that knows itself to be a large social consensus) against their claims. Another is that large and complex problems whose major effects are in the future and whose solutions require consistent long-term global cooperation, and would require powerful entrenched interests to alter their behavior or even give up power and profits, will always be susceptible to efforts by those interests to obfuscate the question and deny the reality of the problem. A third possible problem is that huge amounts of social and political energy will be wasted in acrimonious, unresolvable arguments that, because they cannot be reliably moved from thesis vs. anti-thesis to accepted common synthesis, will (given the kind of creatures we are) devolve into polarized tribes of competing claimants who have no productive route forward and nothing more effective to do to win mindshare and cultural dominance than to get louder and more angry, and, possibly, violent. When arguments cannot be resolved to most reasonable people's satisfaction one way or the other, everything becomes a religious war of fiat against fiat, and everyone (including in the end even unscrupulous actors and powerful entrenched interests) loses because the world - that single dim, flickering spark of self-aware creatures in an unthinkably huge expanse of empty physical order - which might have devoted its collective energy to vast expansions of global and even interplanetary wealth, well-being, and unimagined human opportunity, didn't.
From speaking abstractly, to speaking concretely: there are a number of problems of unprecedented scale that are either present now, or are on the 10 - 30 year horizon, that I'm deeply afraid existing political and cultural structures are not capable of handling safely. We have managed for ~70 years to not start a nuclear war; growing nationalist sentiment, increasing numbers of increasingly unreliable state actors with the capability, and the increasing tension that the following problems will all bring with them make our continuing ability to walk that tightrope questionable. Automation and increasingly capable artificial intelligence systems will almost certainly take huge numbers of jobs away from humans while also taking away most of the jobs that in any other economic shift those humans would have moved into; massive unemployment and gross inequality of income is, without sweeping and radical changes to cultural institutions, a huge powderkeg of human unrest, of the sort that has driven most historical revolutions and riots. Continuing and increasing effects of global climate change, including sea level rises in coastal cities and areas, droughts and famines, desertification, disruptions to fresh water sources and patterns, increasing serious adverse weather events, altered productivity or viability of certain food crops in certain areas, the spread of tropical disease vectors to wider swaths of previously temperate areas, and, above all, the huge and sudden increases of human migration that all this will cause. The increasing power and decreasing cost of sweeping bioengineering abilities that may place it in the power of more and more states, groups, and even individuals to fiddle with massively complex biological systems, with possibly huge unexpected consequences, not to mention the opportunity for disaffected groups (of whom (see preceding sentences) there will plausibly be unprecedented numbers for a variety of reasons) to deliberately engineer plagues and diseases. The chance - possibly an outside chance - that our explorations into making thinking machines will succeed too well, and not cautiously enough, and we will introduce into the world entities that are at least in some ways smarter than us and who have interests that don't align with ours.
As much as Elon Musk and others are building deeply exciting new possibilities and solving serious existing problems, the near future is - can be - a terrifying place to contemplate. And as much as I admire Musk's desire to intervene in those parts of possibility space that seem to him most likely to have the largest effects on future human opportunity - and even as much as I agree with him about the importance and value of the interventions he's chosen to pursue - it does seem to me that the significant moves most concerned people are making are orthogonal to the real problem. The real problem seems to me to be political and social: we have, and have had for a long time, the technological tools to solve basic human physical needs like water, food, and shelter, and even personal and social needs like education, the opportunity to pursue one's own vision of happiness, and social connection and dignity. But we have not had, and still do not have, the insight or cohesion to find ways to move collectively forward cooperatively, proactively, and without violence. If the end of the human race comes to us, it seems to me far, far more likely that we bring it on ourselves through shortsighted selfishness than that it come unexpectedly from without. The root of our problems is us, and our (in)ability to coordinate, cooperate, and agree. No technology, if it doesn't address that, can do anything beside tangentially touch that core problem.
There is a coordination problem at the heart of humanity's millennia-old struggle to build enduring civilization and lasting peace. We need a coordination solution. This fundamentally requires a deep and public clarity over which claims about reality are viable, which are not, the relative strengths of each, and the complete following out of all the possible arguments involved. I argue that this is, despite all appearances, doable.