Don’t get me wrong. I still want to write. And I still will. For myself, such as in my journal. And for my work, which provides plenty of opportunity. And for my blog, when I have something valuable to share. Just less for my blog. At least for now.

Here are some reasons why:

• I’m not sure it adds value to the world. It ends up being a lot of opinion. I don’t know if the world needs more opinion. The world needs more facts. And more awareness of and understanding of many known facts.
• It encourages unhealthy behaviour. I don’t want to be thinking about my blog when I’m trying to be in the moment. I want to be present in the moment as much as possible. I don’t want to put undue pressure on myself that causes unnecessary stress. I don’t want to write something just to maintain a streak.
• Most people who liked my posts or followed me have suspect motivations. You can tell by the name of their accounts. They’re optimizing for search rank on terms like “internet marketing.” I did connect with a few great likeminded people, but they were the minority. It doesn’t feel like an effective channel for making meaningful connections.
  

So, what’s the plan? I’m not sure yet. I enjoy writing. I like publishing to the world. But I want it to be useful. Meaningful. Aligned with my beliefs and values.

I’m going to think about it. If I have something valuable to share on my blog, I will share it. If I have the interest in and opportunity to publish elsewhere, I may do that too.

It was a worthwhile experiment. I learned a lot. Now I want to build on that learning in a way that’s right for me.

The video describes research by OpenAI on AI agents playing hide-and-seek. There are hiders, seekers, and an environment. The agents’ reward function is simple. Hiders get rewarded for staying hidden. Seekers get rewarded for finding them. Variables control how they can hide, seek, and interact with the environment.

After millions of rounds, the agents learn how to hide and seek. They learn how to cooperate. And they learn how to cheat by exploiting a bug. The bug allows seekers to “box surf” into a hider refuge and find them. (The video linked above shows how this works.)

Such cheating is so common in reinforcement learning, we should expect it. Give AI a goal, and it will optimize towards achieving it. In doing so, it will exploit anything it can. It will find ways to achieve a goal that we never thought of. These are often ways that we would consider cheating. Take a look at this list on Boing Boing. My favourite:

In an artificial life simulation where survival required energy but giving birth had no energy cost, one species evolved a sedentary lifestyle that consisted mostly of mating in order to produce new children which could be eaten (or used as mates to produce more edible children).

This is what people like Nick Bostrom (see his book Superintelligence) fear most with AI. The public may think about Terminators. And we should put limits on autonomous weapons. But a more general and worrying risk of AI is that it will optimize to achieve goals in ways we don’t expect. And some of these may be negative. Bostrom’s famous example? Task an AI with making paperclips and it might turn the world into office supplies.

Seeing AI learn to play complex games, with associated emergent behaviour, is astounding. This technology has incredible potential. But we must figure out how to ensure AIs’ actions uphold human values. Until then, we should be careful about unleashing them from video games.

From Christian’s perspective, humanity’s evolution isn’t discontinuous. All organisms use information to harness energy. Humans are better at it because we can store and share more complex information. This allows us to accumulate knowledge, and then apply it. The history of evolution is a history of storing and acting on information to harness energy. Humans are no exception.

In the book, Christian also shows how the seeds of human development are latent. Take foragers, for example. They had plant knowledge that became useful for agriculture when conditions were right.

When you combine these ideas, it provides a framework to predict humanity’s future. (In broad strokes.) We develop by using information to harness energy. So in future, we’ll need to collect and share more information to harness more energy. The seeds to achieve this are present today, like foragers’ knowledge of plants. The question is: which seeds will bear fruit?

I’ll admit this isn’t a novel insight. Ray Kurzweil, for example, describes evolution as exponential improvement in information processing. But, if true, it’s powerful in its simplicity. We don’t know the exact mechanisms. But in future, we’ll expand our knowledge, and our energy sources. And we’ll use knowledge present today, which may be arcane and hidden, to do it.

At that moment, I vowed to keep track. I created a life expectancy calculator in Google Sheets. It tells me my current age to three decimal places (42.381). It also tells how much longer I may live (35.919 years). For the latter, I’m conservative. I use the lifespan for the average Canadian male my age. I don’t factor my life extension interventions nor my cryonics membership.

Every day, when I write in my journal, I note these numbers. I even use my current age for my journal entry titles. This helps focus my mind. If I have only 35.9 years left, that’s about 35 more of my favourite annual events. Like the first snowfall of the year, and the bloom of spring flowers. (“The Tail End” by Tim Urban goes into many more examples.)

Life is precious. Time is fleeting. It’s easy to get distracted. Stare into the abyss every day by calculating your remaining time. It will help you be mindful and focus on what’s most important.

Life shouldn’t be different. We have many key metrics. Income, expense, and health metrics, for example. But is there one metric that matters most? A north star metric for life?

For me right now, my answer is “mindful moments.” As Buddhist monk and peace activist Thich Nhat Hanh says, “Only this actual moment is life.” So the more moments I’m mindful, the more moments I’m alive in full.

Furthermore, my ability to be mindful is a good gauge of my functioning in all areas of life. If I’m struggling to be mindful, I may not have slept well. Or I may be over-scheduling or over-tasking myself.

When all aspects of my life are functioning well, I can achieve more mindful moments with ease. Hence my number of mindful moments is a useful north star metric. When it’s going down, I know something’s wrong.

Would mindful moments work as your north star metric? If not, what would you use in its place?

I intended to follow that paragraph with a list of the top things that I (and you) can do to fight climate change. I often use the Pareto principle (AKA 80/20 rule) to make action plans. I had intended to do so here. I planned to find the 20% of things I do that contribute 80% of my carbon emissions. I could then find ways to reduce or eliminate them.

Alas, not so easy, as this image from the Prairie Climate Centre shows:

Try to apply the 80/20 rule here, and you see that citizens have little direct control over most emissions. Turn down the thermostat at home? Okay, but that affects 6% of total emissions. Stop flying? Okay, but that affects 2% of total emissions. (This chart shows 1% as it doesn’t include flights outside of Canada. See this global data on aviation.)

Your biggest lever is car and truck transportation, at 19.9% of total emissions. Which is a big chunk. But even if everyone stopped driving, flying, home heating, and home electricity? That’s still only 27.9% of emissions. The balance of 72.1% includes things like oil and natural gas leaks (7.8%) and mining, oil, and gas exploration (14.5%). As a citizen, I can indirectly influence these through consumption. But it’s hard for me to directly affect them.

That’s why political action is so important. So in the end my list for today has two items. First, take the direct actions you can. Drive less, use less energy at home, and reduce consumption of carbon-intensive products. Second, pressure politicians to regulate and reduce the emissions beyond your direct control.

That’s what we’ll be doing in a few weeks at the Climate Strike.

Today is September 14. Yesterday marked 30 consecutive days of posting. Today makes it one month. So what did I learn?

Here are a few notes for others that might consider a similar challenge:

• Use a good tool. I use Ulysses. It is phenomenal for taking and organizing notes, writing, and publishing. (I intend to write in future about how I use Ulysses.)
• Write down ideas when they come to you. Knowing my need for a post a day, I thought of ideas for posts often. If I didn’t write them down, I often forgot them. Having a bucket of ideas to choose from is easier than facing a blank sheet. There are (likely, I haven’t calculated) an infinite number of things to write about. When you have a bucket of ideas, it becomes a choice about what to write when.
• Read, be mindful, be curious, and be well-rested to get ideas. I noticed that I came up with more ideas when I was reading interesting books. Also, when I was mindful and curious about the world. I also noticed that when I’m tired and haven’t slept well, my ideation suffers. It’s harder to think when you’re sleep-deprived.
• Choose topics you want to learn more about. This experiment prompted me to learn more about things I might not have. Like pumpkin spice lattes. It’s an opportunity to research topics you might only think about in passing.
• Don’t set word count goals. It’s daunting enough to commit to a post a day. It becomes much harder if you feel compelled to write a certain length. Write as much as you feel you can each day. Pushing for a specific word count is a different challenge.
• Use your writing to better understand your interests. I decided I wasn’t going to focus on specific topics. I was going to let my interests dictate my posts. I can now see from work over 30 days what topics most interested me. This meta-view provides interesting self-insight.
• Do what you can under the circumstances, and plan ahead. I wrote on planes, trains, and buses. (Not cars. I get sick.) On family vacations and business trips. I factored my need to write a post into my schedule. This allowed me to think in advance about when I might write, and about what.
• Focus on the intrinsic value, not extrinsic rewards. This is hard, but important. Especially if you’re doing this for yourself and not for money or reputation. It’s easy to get sucked into optimizing for likes, follows, and other extrinsic rewards. It feels like a game. But this will skew your posts towards those external factors. You will start to write to get likes and follows, not for its own sake. Try to do it for yourself. It’s easier said than done, but the next point can help.
• Focus on feedback that supports your intrinsic motivation. I do look at likes and follows. But only once per day, after I write my post for the day. I try to use likes and follows as feedback to support my intrinsic motivation. For example, I like to know when someone read something I wrote. It feels different to write for someone else rather than yourself. That’s motivating. I’m also curious about what topics of interest to me are of interest to other people. It’s a way to gauge resonance with a broader community. I also enjoy engaging in meaningful dialogue through comments. This can help me deepen my thinking on a topic. I do find myself drawn sometimes to the dark side. For example, after seeing lots of likes for parenting posts, it can spark desire to write more of them. But I won’t let myself go down that path. It will kill the joy of writing for writing’s sake. I still struggle with balancing intrinsic and extrinsic motivation, to be honest, and hope to better resolve this in future.
  

So what now? I’m going to keep up the pace and keep learning from my experience. It feels like a habit now. And I’m starting to develop a system and process for turning ideas into posts. I’ll continue to share what I learn for anyone who’s interested.

Pumpkin spice is one of those things that feels timeless. Like apple cider and hot chocolate. So I was interested to learn that Starbucks invented the taste and concept in 2003. Wikipedia has a great article on the history of the pumpkin spice latte. It’s now Starbucks’ best-selling seasonal drink. They sold 200 million from 2003 through 2015.

Their success attracted copy cats. This includes other coffee chains. But also manufacturers of unrelated products, like air fresheners. You can see this in Google Trends. Interest in pumpkin spice increased almost 6x from 2004 to 2015. (Google doesn’t provide data from before 2004.)

It makes me wonder what other timeless cultural artifacts were marketing inventions. Santa Claus as we know him, which Coca-Cola invented, is another example. The tradition of diamond engagement rings, which De Beers invented, is yet another. But which traditions rooted in marketing are so old, we’ve lost their origins?

Not that this is necessarily bad. (Though I’d argue that the diamond engagement ring tradition has been bad for men, miners, and the environment.) I enjoy pumpkin spice lattes whether Starbucks makes them, another company makes them, or I make them. But it is good to know where our culture comes from. And how fast we can turn something new into an annual tradition.

Here’s my thinking. Often in meetings, much conversation contains no relevant information. This is because attendees often:

1. Repeat themselves, or repeat other people
2. Go off topic and have tangential conversations
3. Break, such as for lunch, coffee, or the bathroom
4. Use stop words and filler content, and otherwise don’t get to the point
5. Posture and signal status rather than talking substance
   

My guess is that, using the Pareto principle, 80% of a meeting’s value comes from 20% of the content. If true, then a compression algorithm addressing the above could deliver 80% of a meeting’s value in 20% of the time.

I have some ideas for how to do this. You could train a machine learning model to identify 3, 4, and 5. You could use statistical techniques to deduplicate repetition for item 1. And you could use natural language processing for number 2. (You could use it to understand the core topic of a meeting, and when people strayed outside of it.)

Granted, you couldn’t compress a meeting while participating in it. This would only work for after-the-fact analysis. But you could use a system to flag items 1-5 during the meeting itself, and keep people on track.

Let me know if you’re aware of such a product. Or think about how to build it in your next meeting that could use some compression.

Contrast the freedom I had with that of children today. In some states (for example, Florida), the police would arrest my parents for negligence. Yet the world is safer than ever. The crime rate in Canada has dropped since my childhood. In 1983, there were 8,470 police-reported crimes per 100,000 people. Last year, there were 5,489. The US has seen a steep drop in crime too.

I thought about this after reading The Coddling of the American Mind. The book looks at what’s undermining free speech and free thinking in US universities. One is a culture of safetyism. Kids need freedom, including freedom to make mistakes, to develop. By over-protecting kids, parents put their kids at future risk of being unable to cope with the real world.

Yet while I appreciate this perspective, it’s hard when it’s your own kids. There’s no clear line between protective and overprotective. Also, while crime may be dropping, pedestrian deaths by car are increasing.

So my takeaway is this: by default, let my kids roam free. When I feel an urge to protect them, I can check that urge against the stats. Based on that criteria, I’m not that worried about my kids playing in a park by themselves, as I did at their age. It’s crossing the road to get there that scares me.