Featured Post

Rest

 I hope that everybody in the world gets their infinite moment of respite today. 

Thursday, March 28, 2024

AI economics

 "Money" is just a proxy for resources. 

"Jobs" is just a proxy for labor resources.

The resources and labor exist whether money or jobs exist or not. It is the distribution of said resources that necessitates a system of money and jobs. 

We should not confuse the measurement of something with the thing itself. Measurement stands only as proxy and when the concept fails, we should fall back on the ground truth.

For instance, for many, the fear of AI is based around the fact that it will automate away many jobs. But the focus on "jobs" as a measure of economic health and prosperity only holds under certain circumstances. And said circumstances may not hold under an AI-driven economy. The same human resources exist whether AI is adopted or not. So the real problem, which people inherently understand but don't necessarily express correctly, is that during the transition period AI will make certain skills less valuable to others, and there will be cost for re-skilling people, while at the same time the rest of society will benefit from lower cost. Also relevant is the risk of monopoly as AI companies subsume entire industries.


Tuesday, March 26, 2024

Relatinoship dynamics and "optionality"

Incomplete, just notes:


 https://www.youtube.com/watch?v=lad3sX6cPJo


"facts"
- women leave men at higher rates (2x)

- men mature slower

- men seek younger women regardless of their own age

- men live shorter lives

- age gap M > F


Consider: "viable dating age" for women vs men -- women have shorter timeframes, hence denser "time in a relationship"?

Optionality? Or is the real problem that younger men are competing against older men for a smaller pool of younger women? 

What about the fact that women leave more often, initiating breakups? Simply, competition
Yes, evo bio is a part of it. no, that doesn't make women (or men) "superficial" and self-serving. Women seek a partner who will support/love and is mature. That's all modeling on the micro-scale though. Given a dating time distribution, it could describe any kind of micro-level situation. The same picture happens even if men break off the relationship but have trouble finding the right person due to competition. 

How about we enhance our model to talk about the economics of optionality? Do women really have more options? Supply vs demand? 

In general, my goals are:

-Model the econometrics of dating such that it may possible explain popular observed phenomena such as "men are terrible" and "women have optionality" as either secondary inevitabilities of the mathematics, or more primarily motivated by the mathematics by some causal mechanism.  

    - Do women actually have optionality? In what sense?

    - Are men incentivized to be more awful? Are they? Or is there an inevitable reason why that sense would arise/become popular? 

    - Does this model give us any other insights? Explain other phenomena? (Crazy cat ladies, e.g. women left behind by the dating market, their number, divorce rates, etc)

https://medium.com/the-renaissance-economist/how-an-economist-sees-the-dating-market-eb3afe0c9e35

Tuesday, December 26, 2023

Ideas and keys

 Effects of disfluency on cognitive and metacognitive processes and outcomes | Metacognition and Learning (springer.com)

Lacan was purposefully obscure so that his readers might work at the meaning and some may arrive at the right one. Maybe there's something to that. Ideas, after all, seldom translate directly into language. So to avoid the audience from attempting to interpret directly, we obfuscate, communicating that work is needed to understand. 


Book Review: Sadly, Porn - by Scott Alexander (astralcodexten.com) 

Thursday, October 5, 2023

Social capital and societies in low-trust equilibria

 I came across a really interesting Reddit thread today.

(1) India is a low trust country, but we aren't even aware of it. : india (reddit.com)

It lists various reasons why India is in low-trust equilibrium. It made me wonder if the U.S. was heading in that direction, or other countries, or the world, for that matter.

Social capital seems to be a central factor in peoples' well-being in a country, regardless of wealth or advancement. I'm sure that there's other factors and maybe social capital is more of a dependent variable in service to some larger picture, but ceteris paribus it seems to strongly correlate with lower crime and healthier institutions. 

I think social capital is one of those things that affects individuals strongly -- it reflects on their attitude and expectations about how others will behave. People from low-trust countries such as India will literally trust strangers much less than people, say, from Denmark. 

Sunday, September 24, 2023

Fungibility of progress

 Building off of my previous post regarding the fungibility of experience. As it turns out, even progress is somewhat fungible. As long as you're making progress on something, you're getting somewhere.

It's like this with lots of complex systems -- if there are enough paths and interconnections, pretty much any path you choose, as long as it's in the right direction, is about as good as any other.

There are limitations, big caveats to this, of course. For instance, you can waste your time on a lot of busywork. A lot of people do this because they can only see the short term, and neglect the big life-changing tasks that could help make your other work a lot easier. Poverty can do this to you, as does anything that induces a scarcity mindset. That's not even mentioning the fact that often poverty actually forces your priorities in the sense that you really don't have the resources and time to create value, so it's not purely a mindset thing.

At the same time, this fungibility framework helps you overcome that. If you're afraid of neglecting some menial task, you don't have to be. As weird as it sounds, keep in mind that in a way, you're overall making good progress, even better progress than if you hadn't neglected that task, so it's not such a big deal. I'm not saying neglect to shower and devote that time to projects, but it really puts things into perspective about what's important and what's not, what you should prioritize with your time (you can't do it all). 
It's just easier this way, because it allows you to assign value-per-unit-time to your tasks without worrying about all the particular outcomes of all your particular tasks. It linearizes life for you: just focusing on maximizing value and you're good.

Again, do take showers and clean and pick up your kids from school and so on. This framework allows you to pick your battles, for those who simply try to fight them all. But some battles, obviously, you just need to fight. 

Fungibility of experience

 People think in careers all the time, but I think from time to time it's more helpful to think in a more "first principles" manner.

 Focusing on a particular discipline or technical skill works for a lot of people when the world needs those skills, and to a larger extent the need is defined by the supply. If you have those skills in your populace, why not use them? 

But things change and there's no way of predicting where the masses are going to go and what particular skills are going to be in demand. At the end of the day, there's no guarantee that your particular skill will fit in with the global "plan". 

What I've seen though is that you can't really go wrong if you just work on generally challenging and developing yourself and being active. People who work like this tend to even be at the forefront of change. 

  • After all, there were no M.S.'s in Data Science before data science became a thing. When data science was being in its infancy it was just smart people applying computational models to business data. It was experts from different fields coming together and defining the practice
  • From the perspective of someone who wants to hire people to work for them, it's good to have specialists who are extremely good at what they do, but otherwise or in lieu of that you just want smart, well-rounded people who can problem solve and are a pleasure to work with, people you can communicate with and will understand your vision
At the end of the day, you can practically work on anything to develop those skills. It's what people like to call "soft skills". It's about developing yourself as a human. It's those skills that are useful in any sort of work: intelligence (emotional and cognitive), communication, critical thinking. 

There are also skills you develop that aren't as general, but generalizable -- developing yourself as a knowledge base. For example, if you like doing Raspberry Pi projects in your spare time, it's likely that you're going to be able to apply that to a business. Even if you can't directly use it, you'll likely find it easier to learn more technical subjects because of the similar "thought-patterns"*. You're building up patterns of thinking that can be easily applied elsewhere. 

And statistically speaking, as long as that base of knowledge upwards, you're likely going to be able to apply it somewhere. For instance, as a child you're impressed that your parents can seemingly do everthing -- and that's because they have years of this thing called "experience" on you. This is what I call the "fungibility of experience". It happens because our world is vast and interconnected -- really, any experience is useful somewhere, someday. 

Often, it's less important what you do, and more important that you do -- maximize "resource production" per given time, even though these "resources" don't seem so "fungible" at the time. 

So based off of this, this would be my "career advice" to myself: 
  • Specializing isn't bad. Specializing builds character, intelligence, complexity. So invest time in learning something deeply, but as you progress, keep an eye on how much you are developing as a human and as a knowledge base as opposed to how much you are simply advancing in that particular subject. 
    • Practical example: If you're a engineer, pick up a book on, I don't know, anthropology. Read different stuff. Learn an instrument, I don't know. More engineering makes you better at engineering, but has diminishing returns on making you a better human. 
  • Generalizing is good. Don't be afraid of picking up new things (and dropping them) to the extent that they provide you with human and knowledge enrichment, and deeply enough where you're actually building some complexity and knowledge rather than dabbling. Connect with others, fast-track your knowledge rather than discovering everything yourself. Don't worry about if it's a waste of time, because again, fungibility of experience.  
    • Practical example: Read research papers in fields you've never even heard of. You'd be surprised at the novelty -- there is a lot of profound stuff on things you never even gave a thought to, if you go high-level enough
  • Be open to a wide range of opportunities. Experiential fungibility only happens in a sufficiently large system, and if you're closed off to opportunities, you're cloistering yourself in a smaller subsystem. Don't limit your opportunities to, say, "environmental psychology" because that's what you got you Ph.D. in. Labels are just labels. 
    • Practical example: Talk to people. What are their needs? Keep up with the news, but not just the mainstream. Where do you think you can help? 
  • If nothing else, focus on simply doing, and more importantly, trying. Maximize that vs. the time you have. Remember, it's not the fungibility of time, it's the fungibility of experience. Just because you've been around longer doesn't make you any wiser. 


* It's what people call "transferable skills", basically -- I just don't like thinking about skills that exist in domains and transferring, but skills that exist in their own right, in a platonic sense if you will, outside of any particular domain of application. It's also imprecise because you could either be talking about skills that are directly applicable in say, a new job, or skills that aren't directly usable, but help you learn your new job by analogy. People actually have been using the term "mental model" a lot more now, which is pretty much what I'm talking about.