I posted to HN, and then my website was flooded with AI bot traffic

The goal is to give something that's still enjoyable to work with, but lets you really nail down the behavior when you want to (and still keeps it ergonomic and fun).

A secondary goal was to make something that is reliable and stable, and is (hopefully) easy to adopt in existing codebases. check it out!

if ur reading this i can publish to my blog w/rss.

im using org-rss-export-to-rss with some very simiple tooling. i thought about using hugo but idk it seemed too heavy and i kinda like the minimalist feel. im optimistic that hugo will be better when i set it up right, and i will probably need some of its features eventually, but thats a battle for another day.

I played my first game of Wordle this week, and I got in 4!

Ok, I got a little bit lucky, but it wasn't all luck. While I was playing I had a chance to think through some of the strategies, and saw some pitfalls I think others might have missed.

Unless you're trying to make your final guess to win the game, you shouldn't guess the same letter in the same place. You already tried that, so doing it again doesn't give you new information.

Your "guess" is guaranteed to be wrong, but you now rule out (or find!) another possible letter when you go to actually solve the puzzle. Lets see an example:

Suppose you have B - A - _ - E - S, all green. You just need to guess the last letter, but there are a lot of remaining options. Among letters you haven't guessed in that slot, D, K, N, R, S all work. You could guess every letter individually, but you most likely won't get it first try, and worst case it will take you five turns. Instead, if you guess D - R - I - N - K, a word that you know is wrong you "guess" every letter at once, so you're guaranteed to get it in 2 (you guess S if there are no yellows).

Just for a moment, think of worlde a little bit like binary search. Your goal is to learn as much information as possible with every guess. If your guess splits the problem space in half (i.e. you eliminate half of possible words if it shows up green, and half of words if it shows up yellow), you learn (on average) the most possible information with every guess.

Now – you only get to guess a letter in one position, so either it's in the one specific position you chose, or it's in any of the other possible four positions. If you want these two outcomes to be roughly equally likely (or as close to equally likely as possible), then that one position you chose had better be really likely to make up for 4 other slots.

You learn (on average) way more about the letter 'Y' with a guess like R - A - I - N - Y, than you do with a guess like Y - E - A - R - N.

Basically the same logic as above. It's either one of the 5 specific letters you guessed, or it's one of the 21 other letters. You want to guess the 5 that most closely split the problem in half.

Going a step further, once you have a few lines guessed, you want to guess the most common/most evenly dividing letters given the letters you already have. Q is a bad first guess, but if you already have a U down, then maybe…

Remember, your goal is to "divide in half" with every letter. If you have a sense of/random sample of the words in the space, it's easier to pick the letters/positions that divide that list the most precisely.

Charles Duhigg was a reporter for the New York Times, and you can feel it in his writing. This is a modern-science report that's sensational, inspirational, and often highly misleading.

This book has it all, self-help transformations from couch potatoes to superstars, the success of Starbucks and the Colts football team, all simply explained by the power of habit. Of course habbit only plays a small role in each of these things. Of course every football game comes down to a little bit of luck (even left up to the wind and the weather). But – someone adopted good habits and later become successful – BOOM – causation – done – problem solved.

Beyond the wild conclusions though, there is some real (and recent!) science being done, and our changing understanding of habits is both interesting and powerful. For me the key takeaways are:

Habits circumvent our normal process of reasoning (and aren't stored in the same brain-area as planning and short term memory) Habits can be modelled as a sequence of trigger – action – reward. If you're missing one of these pieces, you won't trigger (or learn the habit loop) Habits can live latent for a long time, even as we move on from them. Make no mistake though, this book is meant to be transformative. What can YOU do to change YOUR life, just like Peyton Manning did? Again, the science is mixed here, but it seems like the author's point is:

Habits are hard to un-learn, but it's easier if you break them down into the trigger – action – reward lifecycle. If you can swap out the action, but keep the trigger and reward the same, you can hijack your bad habits to learn good behavior. If you want to make a transformative change, it's easiest to focus all your attention on one habit, and then the rest are likely to follow. Habits can be a good thing. Thinking is hard, and deciding what to do takes effort. If something (like choosing what to make for dinner) takes mental effort, your less likely to do it, because that mental effort is as much work as the actual cooking. Try to form habits, and try to think about what you're going to do ahead of time, so that when you get to the moment of truth, it doesn't take effort.

Like any political agenda, modern monetary theory isn't perfectly specified, and people often tailor its definition to their own ambitions. As best I can tell it goes like this:

If you're a country that controls its own currency, its ok to go into debt, and you can print as much money as you want as long as it doesn't cause inflation.

Some people argue that with the power of MMT, we have the money to finance all of our wildest hopes and dreams, if somebody would just have the courage to turn on the money printer. Free college, loan reimbursement, free housing, we could pay for everything with a click of a button, and no negative long-term reprecussions. As long as we don't cause inflation, there really isn't anything to worry about.

The thing is, that's a pretty big if. You can "prove" anything you want if you start with a false assumption. If we could print infinite money, and not cause inflation, that'd be pretty awesome, but is such a thing even possible? How did we get here, and why does anybody believe we can do this?

In theory, economics is based on supply and demand. If everyone wants apples, but there are no apples available, the price of apples goes up. If nobody wants apples, but every citizen has a stockpile of 100 apples, the price (and the implied value of apples) goes down.

In the same way, if everyone in the United States magically had 12 million dollars, they'd all be competing to use that money to buy the same limited supply of things. Many people want 3-bedroom houses, but there aren't magically more of them. It's the same people competing with each other to buy the same house. In order to outbid someone with 12 million dollars for something they really want, you also need to spend millions of dollars.

Fundamentally, people want money so that they can use it to buy goods and services. On a global scale, if every government printed more money, but didn't also increase the global amount of available goods and services, you're not increasing everybody's real buying power, because everyone is still competing to buy the same limited number of goods and services. You only increase real wealth if the number of goods and services produced also increases.

So if America went out tommorrow, and started printing tons of money, and taking on tons of debt, we'd expect the cost of everything to go way up. We'd expect government bonds to be much less valuable, right? Right?

Well it turns out that during the 2008 financial crisis, the U.S. did print a lot of money, and did take on a lot of extra debt, and everybody expected inflation to happen, but somehow it just… didn't. Markets are hard to reason about, and nobody really understands why it happened, but people have thrown around some theories:

The United States is considered a trustworthy institution, and people need a safe place to put their money. The world financial system is already so invested in the U.S. dollar that the U.S. can do whatever it wants. The stimulus bill spurred so much economic growth that it offset inflation.

Ok, so, we took on a lot of debt before, and it was fine. We printed a truly wild amount of money, and people still trusted us as a stable, reliable institution. Everyone believed that the U.S. government would pay them back, and would pay them back in dollars that held value.

Even in a crisis caused by the U.S. economy, the U.S. looked like a safe place to put your money because the U.S. always had been so responsible, and so reliable in the past.

So – if we suddenly started being irresponsible, like, racking up huge debts and printing money to pay them back, people will keep lending to us forever, right? Right?? If we don't produce anything, we can just keep printing money forever, and people will just give us stuff, right?

There's this idea that because we're borrowing in U.S. dollars, and that U.S. dollars are the de-facto currency of the world, we're never in trouble, because we can just print more money to cover our losses. But do you think people will keep lending to us if, in terms of real physical goods, we're not offering anything meaningful in return? Do you think people will continue to use USD as their default currency if the U.S. government buys all your stuff, but gives you paper, and no items in return?

We could technically pay off our debts by printing a shit ton of money, but from an investing perspective, it would be essentially the same as default. We made a promise, and in terms of goods and services, we're not providing any real value to fulfill that promise.

Defaulting on a loan is so dangerous because it destroys your credit. If you go bankrupt, it's super hard to get another loan because people don't trust you anymore.

In the same way, even if the U.S. doesn't technically default on its debt, if it covers up by printing more money, it's essentially going to have the same effect because people won't trust us anymore.

We can "cash in" on our current trust and influence, but once we break that trust we can't rely on it as strongly anymore.