Empathy and Invisibilia

The new episode of Invisibilia is at times both incredible but ultimately, frustrating to me.

It starts with an interesting premise: a new producer, Lina Misitzis, is given the task of cutting a raw interview into a small story as a test, but the hosts are shocked at the difference in tone of her work when compared to their own version of the story they produced concurrently.

Much credit to the host, Hanna Rosin, for having the humility and courage to consider the possibility that maybe her existing team got their version of the story wrong, and allowing for the question of whether or not Misitzis’ story got closer to the truth.

From there, they play NPR’s full story about a sympathetic interview with a former “incel” and the moment he realized he could change. It’s a classic redemption story, much like the ones I’ve heard about ex-Klan members and those that left cults or restrictive churches before on NPR.

Then they play the full version of the new producer’s story, and it tells a different tale. She doesn’t cut out parts of interviews that cast the subject in a bad light, and those same interview clips take on entirely new meanings when a fuller picture emerges.

Rosin then goes all-in on how NPR hosts typically approach stories with unlimited empathy, how putting the listeners and viewers into the shoes of their subjects is their bread and butter. She brings on an expert to talk about the importance of empathy, and disturbing trends in empathy decreases in younger people, and how a world without empathy would be a terrible place.

It’s at this point that I started to get a little upset about this episode. It quickly felt like a Boomer/GenX vs. Millennials moment where younger people are being blamed for being more reserved or careful with doling out their empathy. It feels like Rosin sets this up as a strawman argument, where if you can’t have total empathy you must have none, and that younger generations were well on their way to ruining society by making it free of empathy.

I wish the show recast its glare on the story prepared by existing NPR staff. Why did they cut out parts of the interviews where the subject downplays the feelings of his ex and underplays the danger he was putting her in? He says a lot of terrible things that ended up on the cutting room floor because it didn’t fit the narrative. To me, that’s the real story the new producer uncovered.

Why can’t the show’s hosts talk about a middle ground—one that I myself have come around to—which is offering strangers the benefit of the doubt and full empathy as a default, but also knowing when you are met with someone truly toxic (like the subject of the episode’s story) that it’s quite alright and downright healthy to completely cut off, shut down, and/or avoid them. Instead, they go further on a tangent about how being selective with your empathy (like people do with politics) leads to tribalism, but honestly it sounds childish and ridiculous.

The more the episode progressed, the more I was in support of Lina Misitzis’ take. The subject of their interviews is a truly awful, abusive person who committed many crimes that he consistently downplayed. He hadn’t changed his retrograde views on women at all, his story wasn’t a classic redemption story, and NPR’s traditional coverage style was sugar coating it and in the wrong.

At the very end, they give the show’s subject one last chance to come clean, with a quick interview featuring hosts of both versions of his story, and he makes it perfectly clear he didn’t change much, that he was just trying to repair the damage his actions caused in his own life, and that he’s only interested in making life easier for himself. It ends on this discordant note.

I love Invisibilia and I love the broad range of stories in past episodes, but the way Rosin and staff treated this one feels really off, leaving me feeling uneasy and dissatisfied with lingering questions they never fully answered.

Woodburn Oregon Tulip Festival 2019

This morning I got up around 5:30AM to get to the Woodburn Tulip Festival around 7AM to try and catch some sun between weeks of rain while the tulips were near their peak in the fields. It was muddy and I was slipping like crazy but I really enjoyed the Spring color on display.

The tulip farm is wildly busy during days when the fields are near their peak, but the key to catching good photos there is to buy a Sunrise Pass if you can only go one day very early (when the flowers and weather align) and you can drive in an hour before sunrise (maybe 5AM?) even though the gates officially open at 8 or 9AM depending on the day. Few people know about this early option, so chances are even on perfect mornings you might only share the acres of fields with a couple dozen people.

A glorious pi hole

Real-time stats on ad blocking in the pi-hole web UI

I recently heard about an intriguing project called pi-hole. You use a cheap tiny raspberry pi computer that runs on your local network, and it handles all your DNS. It’s essentially just a database of ad servers that run as a blocklist after grabbing DNS from popular services (like Google, Cloudflare, etc), so you get all the DNS you need on your network, but without any connections to adservers.

Granted, I used to run an ad-supported series of websites, but I tried my hardest to approach it ethically back then (and I subscribe to half a dozen news sites now). It wasn’t until last year that I finally started using adblockers in my browsers, and as the ad blocking wars have escalated, I’ve gotten a little tired of the nag screens and the small news sites that when I whitelist see they’re talking to 77 different servers to serve me up a single page. I understand free content sites have to find ways to make money, but is the answer taking away the privacy of your readers and sharing all their data with dozens of ad companies? Then there are devices I can’t disable ads on, like my TV.

A few friends tried out pi-hole and said it worked pretty well, so I gave it a try myself. I ordered up this complete raspberry pi kit from Amazon, then followed this guide to get things set up and configured. You can skip much of the first half of the guide since the Amazon kit ships with a bootable OS on the memory card. You do need to have a HDMI connected monitor (or TV) and USB mouse and keyboard to set up your raspberry pi once, but once you enable the SSH interface, you’ll never need that monitor or keyboard/mouse again.

Be sure to install the blocklists and whitelists as described in the guide, and don’t forget you can whitelist any future sites using the web front-end to pi-hole, through a simple form.

Getting everything configured

I had to reconfigure my router somewhat, to both give the raspberry pi computer a permanent internal IP address and to divert all my router’s internal DNS to the IP address of the raspberry pi. And if the raspberry pi ever goes down, your DNS goes with it, so you’ll notice immediately.

Once installed and running, the web felt a little snappier in a browser without an ad blocker installed. Most all ads are gone and according to my pi hole stats, 20% of my internet traffic that was spent on talking to ad server requests is freed up. No more whitelisting nag screens from sites, and finally, no more ads on my TV. About the only ads that remain are in the Instagram client (which I would pay Facebook money to remove if I could).

That big tall rectangle on the side is where ads used to show up

It’s not the easiest project to tackle because you have to know your way around your network settings, and it will take about an hour to set up, but once you have it running, it’s pretty great.

American healthcare still sucks

I picked up my prescriptions yesterday and my new health plan gave me a 90 day supply of pills. This is how it used to be, but for the past 6 years my previous plan only gave out pills for 30 days at a time.

Though it sounds minor, it’s huge.

It’s a pain to refill prescriptions. I have to call in my bottle number to a recording, worry about how many refills remain (if it hits zero it adds 1-2 days to refill), then hope the Rx shop has them on hand or can order them before I run out.

All told it’s 2-3 days out of every month of added stress and anxiety about getting new pills and heading into town to get them. So figure 10% of every month, I’m worried about my medications.

When my health plan changed me to 30 day limits they claimed it saved them money, and I guess that’s technically true, but it comes at the cost of a monthly hassle for every single person using the plan.

I’m looking forward to going through this song and dance only four times a year rather than 12. It’s like a weight has lifted.

Stubborn thinking

The start of the Banks-Vernonia rails-to-trail

There’s a glorious plan for a rails-to-trails project in my neck of the woods in Oregon. It’s been planned out for years and funds have been raised. For good projects like this, there are so many available grants (in the millions of dollars) that the county bought up all the necessary land easements and was set to begin construction on the trail soon. In the area, there’s basically no way to get from north to south safely, as the only paved path is a high speed 55mph narrow two-lane road with no shoulder and no bike lane.

The region is economically depressed, as there are no major employers or manufacturers, so pretty much everyone works and commutes elsewhere to larger cities. A rails-to-trail network would be a boon to the area. I saw the transformation myself with the Banks-Vernonia trail about 20 miles north of this proposed route. 15 years ago I traveled to Vernonia for an event, and being a former lumber town, I’ll never forget the day I was there because all the shops and restaurants were closed up and there was no place to get lunch on a Saturday. The place was a ghost town. I ended up getting food on the way home a few miles away. When I returned to ride the actual trail in 2014, I remember it being a tuesday afternoon and when I pulled into Vernonia on my bike, every restaurant was crowded, the streets were filled with people and I ended up getting lunch from a hot dog cart (and even that had a line).

Imagine a railway from the 1800s that has sat dormant since the 1980s, turned into a wide, flat paved off-highway road for families to enjoy while they spend their time (and money) in the region. It’s a no-brainer and a big win-win, right?

The miles of trail that will someday be used by thousands crosses through a couple dozen farms, and this being a freedom-loving conservative area, they’re not fans of any of it. While everyone has had a railroad easement on their property since the 1800s, farmers are always looking for arable land even when it’s only a couple percent of their overall space and even though the county will build fences for protecting their farms (the Banks-Vernonia trail has dozens of custom fences built and paid for by the trail).

The farmers have banded together and lobbed legal challenges at every step of the way, grinding down a lot of progress. Even now that final studies and construction plans are being drafted, they’re turning their opposition to 11. They flood county meetings talking about how trash will line the entire trail and how people running with dogs will somehow frighten or maul their livestock or that the trail will become a home for the homeless. It doesn’t matter that there is zero crime (I was at an earlier meeting where a sheriff from Banks said he’s never been called to the trail and to date had zero crime reports on the trail), no homeless, and little-to-no trash on the Banks-Vernonia trail, or that the only bike trail with homeless encampments is the one in downtown Portland with easy access to city facilities (this trail would be miles from any town where you could even get something to eat).

I really hope a couple dozen farmers don’t win this fight. They’re being extremely myopic as soon there could be thousands of families happily using this safe, off-highway trail to enjoy the region and visit the towns along the way. I have no doubt this will be a huge economic driver for an area big on winery tourism. Having a 20 mile paved path linking the region together safely will showcase everything the area has to offer and thousands of visitors will enjoy it every weekend, most likely stopping at any towns along the way to eat and drink.

If you live in Northwestern Oregon and would someday consider a trip on this completed trail, the county is asking for comments from people in support of the project. The details are all in this Facebook post with a deadline of this Thursday to hear public comment that helps drive it forward.

Optimizing for outrage

I’m dismayed to see Twitter and Quora optimizing for outrage lately in their quests to keep users engaged and using their services at any cost.

Twitter has made two recent changes that prize outrage over user happiness.

The first is inserting recent trending news into my list of notifications. For 12 years, my notifications tab has been filled with nothing but a feed of interactions on my tweets. It’s been 100% personal and customized to just me for all that time. A couple months ago, they started adding in news alerts. At first, they weren’t bad, it was a lot of CNN breaking news level things like “A major congressional hearing is now streaming” and it was kinda useful, but after a couple weeks it quickly devolved into “Trump did something” with a much lower bar. I’ve only recently unblocked Trump (tired of seeing friends RT his ugly messages into my timeline) and I have zero interest in hearing about the garbage things he says each day. If I wanted to know, I’d follow him. I’m sure there are great engagement numbers for these messages, but it’s low information, high outrage fodder to me.

At the very least, give me some control over them, or stick to the expected behavior for the last 12 years.

The second change to Twitter is finally allowing mobile users to sort their timeline by date, like it was in the early days. Last year I quit Twitter and deleted all my tweets, and I stayed away for two months when it became too much of a timesuck. When I returned, one way to keep myself sane was switching to a latest-tweets-first timeline.

The “in case you missed it” feature that is the default sort had grown to constantly stoke outrage. It would show me the most RTed and liked tweets at the top, often up to 24 hours after they were posted. In our current political climate that meant for 24hrs after any major event, I would see tons of tweets stuffed into the top of my timeline, even hours after things were debunked or stories shifted. News didn’t naturally decay as the “in case you missed it” sorting made sure I saw stuff over and over again.

Moving to latest tweets means news comes and goes. Stuff bubbles up for a couple hours and then it’s gone. But the outrageous thing about this feature is Twitter mobile pushes me back into the Top Tweets/in case you missed it sorting once every few weeks, and I have to immediately switch it back to latest first.

I have no doubt that showing just the top outrageous tweets leads to more engagement. If you’re constantly hitting people with outlandish news stories they’ll open the app more often and interact and post about what they think so the cycle continues. Latest stuff means things decay and not a lot dominates your timeline for more than a couple hours.

That Twitter feels compelled to force me into the outrage cycle baffles me. I dislike any time a service makes a change to my preferences for me.

Finally, Quora recently started pitching me on their revenue program. In order to grow their service, they’re now paying people based on questions and answers, and they give me a weekly notification and board of their top earners.

In the abstract, it’s cool they’re paying writers on their platform for their contributions, but when you build in incentives, you create a system of rewards and it’s hard to make sure that system encourages the best kinds of contributions. Ultimately you want to greatly improve your experience as a result of giving people money for their contributions.

Naturally, people like to game these kinds of systems, especially when there’s a financial reward. Going off my normally great and interesting Quora top questions emails, they’ve taken a dark turn in the past few months.

Is it because I click on questions that sound bananas according to Quora’s algorithms so they show me more outrageous things each week? Or is it because people are getting paid to make highly engaging questions that set people off? I can’t tell, but it feels like the service is measurably worse, since all Quora wants to show me are clueless questions from Trump supporters. If you look at the subjects above, it’s almost a form of sealioning as people are “just asking questions” but my weekly digest email feels more like a pit of vipers than ten interesting questions I used to like reading.

Content moderation has no easy answers

This morning I read Casey Newton’s expose of Facebook moderation problems at the Verge.

Let me be clear upfront: content moderation is tough and I have no idea how to solve it at internet scale—in fact I’m not even sure it’s possible to do on the orders of millions and billions of items to be reviewed. Stories like this started coming out about 5 years ago about facebook moderators in the Philippines having high burnout rates and I remember thinking the problem had no easy solution back then (hint: it’s even worse now).

I ran a somewhat popular indie site for 15 years, the last half or so with ample moderation. But to put the scale of the work in perspective, we were dealing with 10-15 thousand active people daily posting about 3,000 things. Slightly big numbers but still small enough you can wrap your head around them. Mostly day to day we broke up bickering matches between two grad students on the site. And even that was still a drag and after many years doing it I had to hang it up to take a break from the day to day stress.

People often say to me that Twitter or Facebook should be more like MetaFilter, but there’s no way the numbers work out. We had 6 people combing through hundreds of reported postings each day. On a scale many orders of magnitude larger, you can’t employ enough moderators to make sure everything gets a check. You can work off just reported stuff and that cuts down your workload, but it’s still a deluge when you’re talking about millions of things per day. How many moderators could even work at Google? Ten thousand? A hundred thousand? A million?

YouTube itself presents a special problem with no easy solution. Every minute of every day, hundreds of hours of video are uploaded to the service. That’s physically impossible for humans to watch it even if you had thousands of content mods working for YT full time around the world.

So everyone says “I guess AI will solve it” but then you have all of AI’s problems on top of it. Baby videos get flagged as porn because there’s too much skin tone filling the screen. Subtle forms of abuse aren’t picked up because the patterns don’t exist yet in the AI and every day is a cat-and-mouse game to stay head of AI. AI is prone to the same biases in the creators and will have negative effects down the line.

I don’t know how to counteract the effects of moderation, or how to mitigate the toll it takes on people. I know this from friends working all over the tech industry, but any job that requires you to solve problems for people and express empathy for them, whether that’s in a chat window or on phone support or at a genius bar, it all takes its toll on people doing it and those jobs have high turnover rates. Many content items described in Casey’s piece are horrific and I don’t know how to you prevent it from harming employees, but aside from those special cases it’s extremely hard to keep the work from grinding people down.

Honestly, I wish there was a solution. I’d love to see Twitter do a better job keeping terrible people off their platform and stopping things like brigading where you make a joke about a public figure and then thousands of people hound you from some unknown source. I wish YouTube would get better at filtering out conspiracy nonsense and stop radicalizing people. I wish Facebook could keep their site free of brutality without permanently harming workers who have to look at it.

I was part of a small corner of the internet where we made it work, but it was downright tiny compared to the big internet scale platforms. That’s not to say it’s impossible so we should throw up our hands and give up, but I just want to acknowledge how hard the problem is to solve. I’ve thought about these issues for decades but there are no easy answers. I don’t let any large platform off the hook for what takes place there, but I do recognize there’s no magic solution.

Two months with Blue Apron

I’m coming to around to meal kits about five years too late, but in an effort to help share the dinner workload at home, around the new year I decided to try out Blue Apron twice a week at first, and now three times a week.

Going into this, I had almost no familiarity with meal kits or Blue Apron, as I’d only heard they’re kinda wasteful in terms of packaging and having to fly food to you overnight. I wasn’t sure if I’d learn much from them, because I figured the ingredients were probably pre-measured and prepped. I wasn’t sure if the service was worth $10 per plate, but my hope was after 6 months or so of Blue Apron that I’d have enough experience working with various proteins, vegetables, and spice combinations that I’d be a better cook in the kitchen. To date, I can only make like 3-4 very basic meals mostly involving BBQing a steak or making pasta.

The one thing that struck me from the start is that the ingredients are all given to you close to the amounts you need, but there’s considerable prep for each meal. I typically spend 15min or so skinning carrots and pressing garlic (there SO MUCH garlic in every dish). For some reason I thought Blue Apron might be Millenial Bachelor Chow with all your stuff pre-cut and pre-measured, but you do have to put in the work to get things ready to cook.

Suggested cooking times are really tight and spot-on. I always choose the middle time in a given range and I’ve had perfect meat and vegetables in every meal. Instructions are pretty straightforward and I’ve only gotten tripped up once when I confused two steps in a meal. I’ve never “ruined” a meal and every dish looks about 98% like the photos that accompany the recipe (the screenshot above is my last 14 meals showing the recipe photo of the meal then a shot of my results next to it.

I’m a bit surprised at how much cooking stuff you really need to do these recipes well. Most use a medium-sized pot and a large frying/sauce pan, but you also need at least a few knives, various strainers, measuring cups, and both a garlic press and a juice press help greatly. Sometimes the instructions would gloss over something fairly difficult to do, like chopping up shallots and I’d have to look up a YouTube video or two to figure out how to do that.

Overall I’ve been quite happy with Blue Apron. About once a week we have a meal that is utterly fantastic, with flavors that are incredible. I’ve taken to calling Blue Apron “culinary training wheels” because I’ve quickly gotten good at prepping and crushing garlic, cutting up and prepping fish and chicken, and chopping vegetables. It really does feel like a quick education in how to prep many different decent meals and I’ve learned a ton in the past two months. Even my spouse who is an expert-level home cook has been impressed with the flavor combos, seasonings and sauces, and overall taste of the dishes.

After two months, it’s been a solid hit. I’m learning how to prep tons of different ingredients and meals and we’re getting really good food that feels like it’s worth $10/plate and would be at home in a restaurant. My original plan was to get enough experience that I could bust out some cookbooks and go to the store to prep my own meals, but it’s incredibly convenient to have it all gathered for you with easy directions. I’ll likely keep doing this long term because it’s pretty fun to prepare, the recipes are easy to follow, and the results are top-notch.

Honestly, I didn’t think it would go this well. This isn’t a sponsored post and I’m not even sticking my affiliate code in a link to sign up. It’s a pretty good service that is both teaching me how to cook as well as letting me make good meals for my family.

A stupid Internet of Things story

Recently my Nest thermostat started getting aggressive about going into Eco or Away mode to save energy. Normally, it’s probably fine but our winter has been a bit chilly this year and I’ve started to notice.

I’ll be working in my home office for a couple hours and suddenly it’s freezing. I’ll check the Nest and it’ll have dropped to 65ºF in Eco mode and will take two hours to get up to 70ºF in the winter. If I don’t get up and visit my kitchen every hour or so, my Nest thinks I’m out of the house.

I realize if the Nest worked with Apple’s Homekit, I could probably figure out a way to use my phone’s location inside the house to denote my presence and keep the heat running normally when I work at home. But Nest is owned by Google, and they’re in an eternal battle with Apple, so that’s not going to happen.

I relayed this nonsense to a friend who came up with a remedy immediately: if I owned a Nest camera, and I put it in my office, it would know I was around and keep the heat on.

He was right of course, but I had to picture it to grasp it fully. In order to keep my house warm when I’m working at home away from my smart thermostat I’d need to buy a $400 camera and point it at myself in my home office so that my thermostat knew I was on the other side of the house.

This is the current state of IoT devices and ecosystems.