>As such, empirical laws that were trustworthy in the past, may become significantly less accurate, and can in fact become dangerous to rely upon too blindly.
Read a really good book on this a few years ago by John Kay and Mervyn King, Radical Uncertainty (https://wwnorton.com/books/9781324004776), where they talk about the issue of applying the logic of "known uncertain" systems (like games of poker, or volcanic eruptions) to completely dynamic "radically uncertain" systems where any attempt to quantify outcomes is effectively completely made up. Election forecasts being one example, where there's no guarantee that you know the rules of the game, if they stay constant between elections and so on.
I think this is very timely because these days you see this kind of cargo cult empiricism where people will say "I am 30% confident that..." as if just talking in a quantitative language means you've somehow understood something about the future or have reason to be more confident.
I've noticed this a lot with climate science, especially. People are waiting for a silver bullet that will reverse climate change and I think in reality, it will have to be a huge, complex, coordinated effort.
Simple, obvious answers appeal to a broader, less-knowledgeable set of people. In fact, it really doesn't matter if there is any basis for the answer, like a lower-dimensional approximation. If repeated enough, more people tend to believe the simple, baseless answer. I feel this is a bigger issue these days than needing to use more complex answers among people who understand the complexity of the problem.
The system dynamics[1] view here would be that complex dynamic system behavior tends to be dominated by feedback loops; and once identified let you ignore many of the other factors. And the trick is to figure out ways to introduce or activate feedback loops that work in your favor, instead of against it. To me that gives some hope of the existence of “simple” solutions to managing complex system behavior over time.
I adore Terence Tao, but this one feels like stating the obvious.
Basically, he says that simple solutions to complex situations work sometimes—for example when a simpler system can approximate your system.
But many times it's not true! So you need a complex solution to a complex situation.
That's kind of stating the obvious, no?
Sadly, while this sentiment may be useful to a small segment of society that is highly skilled in navigating our polluted information environment, the vast majority of our society cannot even maintain a grasp on reality, much less coherently reason about solutions of any kind.
People live in alternative realities depending on if they get their news from CNN, Fox, TikTok, Instagram, X, YouTube, or BlueSky. Here's a recent clip from Joe Rogan that illustrates this perfectly [0]. He talks about the different perspectives of liberals and conservatives on USAID.
It's likely this will get exponentially worse once our social media feeds become full of AI imposters competing with one another for everyone's attention with outrage bait and fake drama and entire fabricated social media bubbles in which each user is made to feel like the ruler of their own little fiefdom.
Video and audio content appealing to every unique combination of biases, superstitions, and political orientation will be available in an endless deluge. I got a preview of this recently when I saw the surreal, gross videos YouTube Reels serves up to my toddler. Similar slop gets eaten up by easy targets on FB. But the slop is going to become more and more sophisticated until everyone is consuming it.
I can only hope the solution comes before it's too late.
If you press people on why they believe what they believe, you will often find that the axioms underlying their reasoning are slogans, trite phrases and hackneyed soundbites.
But here's the catch: the memes, short texts, or short videos are spread through a complex machinery of interrelating parts with global reach. Explanations are plausible, wrong and very profitable in front of the right eyeballs.
aka "reality has a surprising amount of detail"
The elephant in the room is that this is about as excoriating a jab as Terence Tao can dole out against the brute and unconsidered actions of Ol' Musky on our (federal) government.
For our sake, I hope Tao is also correct in that damage to the federal system may not necessarily destabilize structures at other (state, county, municipal) scales as regards socioeconomic organization in the USA.
No, I think he's completely wrong.
The world isn't being ruined by a meme. On the contrary, our problem is the complexity bias, which seems to have developed over the previous century.
As the result, nobody understands anything anymore. I fact it seems that more complex solutions have been worked out, that allow to run on sone kind of protocol, with little to no thinking involved.
Time is being wasted on dealing with problems in overly complex ways, and there is no room left for what can't be simplified.
No offense intended, but this is idealistic and somewhat expected from a person living in an ivory tower.
>long-standing assumptions are tested and updated as new information about the current state of the world comes in
The entire problem is that this mechanism doesn't exist in practice. Real-world systems tend to accumulate incidental complexity that serves no purpose, with no real way to review it, as any potential mechanism is subject to the same accumulation. That's called corruption.
The only way to compress and simplify the outdated complexity that has some chance to work is total system destruction and rebuilding, with previous iterations taken into account. But when you're stuck a local minimum it makes things even worse. Now, if you have another, easier way that would actually work in real world, I'm all ears.
The reason people summarize things, often without being correct, is that the world has become too complicated to model. It’s an outcome of the situation, not just a corresponding trend.
I think Terrance makes a great general point in general, but the specific claim is not always true: complex dynamics can arise from simple initial conditions, therefore, simple changes could resolve those complex dynamics.
A sick burn.
I'd put it this way.
Current society largely (but not completely) relies on experts building highly complex systems. This includes not only public infrastructure like waterworks, buildings and information systems, but also say the very bread you eat (to get it at a low price and high productivity[1]), the computer we're reading this in (mind-numbingly complex system), the internet. It's a feature of capitalism (really, a feature of many administrative systems and product interfaces) that the consumer thankfully doesn't have to be too much of an expert on say computers (say theory of CPU architecture, pipelines, assembly, etc..) to buy a laptop or use software. The consumer only needs to be able to tell which competitor product is best for his use (although often, as he should, relying on expert reviews). We are good at hiding complexity behind interfaces as well, packaging complexity and hiding away its intricate inner workings. All of this enables life in a complex society.
I think it's misguided or hypocritical to completely distrust experts specially when it comes to public policy, public administration and science, given how much we rely on experts for everything else. It's not even much of a choice, I believe: the fact that we rely on those complex mechanisms inevitably will make certain failures that often demand public attention also complex. Say a food company synthesizes a highly complex (and not present in most natural products), but good tasting, substance. Then we kind of need equally complex review of its impacts on human health. A highly complex computer network will need highly knowledgeable (and correspondingly highly complex) solution to certain bugs that might appear, specially in cases like cross-domain failures where complexity encapsulations fail for various reasons. Think how unlikely it would be that every discipline has been exploited to extremely high complexity, but just by chance we could get away with simple solutions for public-facing and public policy problems.
I like simplicity, and I even like the idea I wouldn't often require experts to understand a public or scientific issue of public concern. But I don't think I'm willing to give up most products of complexity, including computers, medical procedures and diagnosis, and more -- and even if I were willing (I might be able to live with say an early 2000s computer :) ), I don't think it's realistic or feasible to really do that. In part because of collective agreement, in part because of for example the sheer population we have to contend with today. Earlier methods of agriculture for example probably can't sustain that many people. We should therefore apply Einstein's wisdom: try to make things as simple as possible... but no simpler. And trust experts when sensible, when the problem at hand is complex enough to be beyond our comprehension (but still important).
Of course, experts can be wrong, but that is something we just have to contend with (like we have to contend with the possibility all the weird procedures we do to produce food or acquire and purify water -- which are managed by experts -- may go wrong, even with significant efforts to otherwise). We can, and probably should, demand explanations (which may be hard to understand for the general public) of the experts and they explain their reasoning. We can examine and expect that their scientific field is healthy, there is consensus and there is a good level of academic integrity. But we should not approach their well informed opinions on important issues from a baseline of arrogance and distrust, because likely they do know much better, in certain cases.
[1] Modern agriculture is highly complex. This includes special seeds, harvesting machines, soil science, weather prediction, and so on. Each of those is in turn highly complex requiring experts to exist at current performance.
Tangential; why is math popular on mastodon like computer science is - but little other subjects? What do physicists use, chemists, mechanical engineers?
I don't like the title. Complex dynamics often require simple solutions. E.g., a complex chaotic system can often be made simple by adding damping (a simple solution).
Because certain concepts have been repeated so often they've become elements of faith (e.g., "taxes are too high", "government is filled with fraud and waste", "taxing the 1% will solve our country's financial problems"), careful conversation about the complexities of the modern world is difficult to have.
I think that complex solutions CAN be grasped by the large majority of voters, but their time and attention are so captured by sloganeers that they aren't exposed to the explanations.