I’m Not OK, You’re Not OK

Paul R. Potts

18 Feb 2021

I’ve been badly distracted from writing in recent weeks, and so have been finding myself staring at an empty text file wondering how to try to catch up and get the writing parts of my brain working again. This time I’m actually doing it under Doctor’s orders – early last week I saw a doctor at Emmaus Health for a check-in on various things, including my mood. Most of my health markers are positive and I am happy about my weight and blood sugar, but I spoke with him about my anxiety and PTSD-like symptoms indicative of trauma. I’m not going to talk about that trauma at length, at the moment, in part because I’ve tried and failed. I haven’t yet finished “processing” it enough to write something decent about it, as opposed to just rambling.

Anyway, the doctor advised me to do, among other things, some of the things I have traditionally done to help myself feel better, including walking and journaling. Grace and I have been getting outside for regular walks. So here comes the “journaling.”

There’s an awful lot going on and not much of it is good. I’m going to try to clear my head of smaller topics before tackling bigger ones. The “smaller topics” include a number of articles that I have saved on my phone to share.

I’m Not OK, You’re Not OK

I felt that this editorial on CNN was helpful and insightful. Thomas Lake writes:

Are you holding it together? I am not.

And

I find myself experiencing this mental-health decline even though I have not fallen gravely ill, or lost my job, or lost my home, or been trapped in a long-term care facility, or endured any of the other major tribulations that currently afflict tens of millions of my fellow Americans. And so I wonder: If the pandemic is this bad for me, how much worse is it for you?

And

One year after the nation’s first confirmed coronavirus infection, there are reasons for optimism. New cases are decreasing. Vaccinations have begun. President Joe Biden has promised 100 million shots in his first 100 days. In the meantime, we must forgive ourselves for doing a little less. It’s OK to soothe yourself with television, Dr. Richards told me. It’s OK not to clean the house today. It’s OK to have breakfast for dinner. It’s OK to not be OK.

Part of the difficulty, for me, is that it is very obvious that a return to normalcy is not in the cards in 2021. In Michigan, it sounds like people in my age group won’t be able to receive vaccinations until late in 2021. I’m trying to come to grips with the fact that we will still be under some sort of restrictions, self-imposed or otherwise, on October 20th, 2021 – the twentieth anniversary of my marriage to Grace.

I’ve been secretly hoping that we might be able to have a modest get-together with friends and family. I guess it could still happen, but we’d better plan an alternative. Besides the wedding anniversary, several of our kids will have to celebrate a second birthday under lockdown, still unable to attend choir rehearsals, or church youth group meetings, or any of the other things that they were doing to form their budding social lives and social skills. That continues to be hard on them, and continues to be something that we can’t entirely fix by taking extra time to check in with them, or arranging some replacement activities via Zoom.

So, I’m giving myself permission to not be OK. And, of course, you have my permission to not be OK, too.

Facebook’s Role in the Capitol Invasion

I want to bring to your attention this article, from the Tech Transparency Project, entitled “Capitol Attack Was Months in the Making on Facebook.” Facebook has made some cosmetic changes, sharing some public relations mea culpas, but it should be clear by now that the company’s leadership is not redeemable and more drastic measures are called for:

For TTP, one of the first signs of mounting danger came from “boogaloo” groups, which we reported in April were using Facebook to prepare for a second civil war, often citing conspiratorial fears about coronavirus lockdowns. Members of private boogaloo groups flagged by TTP later engaged in real or attempted violence—an ominous warning of how online radicalization can spin out of control.

But that was just the beginning. Since last fall, TTP has documented numerous instances of domestic extremists discussing weapons and tactics, coordinating their activities, and spreading calls to overthrow the government on Facebook, up to and including the mob attack on the Capitol, which left at least five people dead. Much of the activity took place in private Facebook groups—insulated communities that allow people to organize out of the public eye while still having access to a large online following.

It’s pretty damning, and makes a very convincing case that Facebook can’t be trusted, given that they have consistently provided private fora where insurrectionists could plot online to overthrow the government via acts of terrorism. It’s isn’t just that they provide an un-policed platform; Facebook isn’t some neutral host of free and open discussion. They routinely disable the accounts of people who criticize the platform itself.

The author of that article was in the same boat that a lot of us are in:

This is a complex situation for me. On the one hand, I favor the eventual dissolution (or at least decentralized transformation) of Facebook. On the other hand, like many people, I received value from using it. And I had intentionally chosen to continue using it for that value despite my concerns.

Personally, I continue to use Facebook only in a very minimal capacity, and a while back I deleted almost my entire history from the platform – at least, as much as I could (I still see the platform recommending that I share old photos which it apparently still has, even though I deleted them). It’s very hard to delete your content from Facebook, and they make it that way deliberately. And the networking value is important for many people, even a barely-connected person like me. What’s the answer? I don’t have a good answer, but strict oversight of all the companies whose business model is to sell your personal information might be a good start.

Gorgon Stare

Thinking about the risks of Facebook leads me to think about, at the risk of repeating myself, the risks of a surveillance state. It was with some interest that I read this review, in Nature, of a book about “Gorgon Stare,” or “wide-area motion imagery.” Per the article,

A camera on steroids, WAMI can capture images of large areas, in some cases an entire city. The technology got its big break after 2003, in the chaotic period following the US-led invasion of Iraq, where home-made bombs – improvised explosive devices (IEDs) – became the leading killer of US and coalition troops. Defence officials began to call for a Manhattan Project to spot and tackle the devices.

Preventing deaths by IED seems like a good thing, but

As the story proceeds, WAMI’s creators start looking for ways to use the battlefield technology at home: having built a new hammer, they search for more nails. Here, the story takes an even more dystopian turn. John Arnold, “a media-shy billionaire,” uses his own money to help secretly deploy a WAMI system to assist the police in tracking suspects in crime-ridden Baltimore, Maryland. Arnold, who has funded other “new crime-fighting technologies,” first learnt about WAMI’s use overseas from a podcast, and decided to debut it stateside. “Even the mayor was kept in the dark,” Holland Michel writes.

In Charles Stross’s Laundry Files novels, he imagines (at least, I hope he only imagines) a technology in which a surveillance camera can be turned from a receiver into a sort of basilisk, turning whatever it views into stone (or, rather, a sort of red-hot charred cinder block). In the novels, that special function of the existing London-area surveillance network is code-named SCORPION STARE. One of the reasons I love Stross’s work is that it extrapolates, but only slightly. And the truth is scarier than his fiction.

A Small Piece of My Past: The TRS-80 Model 100

And now, for something completely different, I want to share an article by Wayne Lorentz about the early days of using electronic devices in journalism. Specifically, the Radio Shack TRS-80 Model 100.

This document exists to show that the urban legend about TRS-80’s being the devices that ushered journalism into the modern age is true. The Model 100 was a workhorse for a number of journalists, including a cadre of roving correspondents for the Associated Press. And also for me: a cub reporter for a radio station in a medium-sized, unremarkable city.

I did not have a Model 100 back when it first came out, but I did get one later. I was interested in using it as a simple, portable writing platform, with a long battery life. (Later, I would try to use the larger models of Apple Newton devices, with a portable keyboard, in a similar way). It was a bit tricky to get text out of the Model 100, in the format that I wanted. At one point, I built a custom RS-232 cable with flow control so that I could upload the text I wrote on a Model 100 to a PC. I also wrote a Visual BASIC program that would receive the text and re-format it with line breaks and word wrapping the way I wanted it.

I didn’t keep it – I just had so much old electronic stuff, and I can’t really be in the business of running a personal computer museum – but it remains an important model for what I would like in a portable device. The Model 100 didn’t really hold enough text to be terribly useful for me, given the way that I write, which has always been best described as “copiously.” It really was better suited to journalism where brevity and meeting deadlines were the major requirements.

In many ways, a modern laptop such as the Lenovo ThinkPad T460s I am using to write this newsletter are just fine for the job. This laptop has a nice, bright screen (brighter and higher-resolution than many other laptops I have owned and own now, including a couple of Hewlett-Packard laptops and a number of Chromebooks). I have a very capable free, open source text editing program, Notepad++, which has most of the functionality I like in BBEdit, my favorite text editor, which only runs on Macs, and I write in Markdown, which is a clean and simple. Markdown meets most of my needs, although I still have more that I want to figure out, such as good ways to tag text for generating indices and tables of contents, and reformat hyperlinks into footnotes when producing printed books. But even this lightweight laptop doesn’t have the battery life I wish it had. And it runs Windows 10, which is constantly harrassing me about signing up for Microsoft cloud services, and keeps trying to make me use Microsoft’s browser, and is full of security holes.

Apple may have pretty much solved the laptop battery life problem with its new M1-based systems, with amazing battery life. But I don’t necessarily want to live in Apple’s ecosystem as well. And I’ve tried many variants of Linux on many laptops over the years; without the proprietary drivers, they often just can’t fully use the hardware features or obtain good battery life. This isn’t the fault of the Linux developers.

It still seems to me that there is room for a lightweight, clean, “just-gets-out-of-your-way” portable platform for writing of all kinds. Maybe at some point I can help develop such a device.

Anyway, if you are at all interested in the history of computers or the history of journalism, I recommend the article linked above.

Farewell to Betsy DeVos

One of the good things about Trump’s defeat is that Betsy DeVos is no longer Secretary of Education. Most people don’t really follow minor public officials all that closely, especially those that keep a fairly low profile, but she was (and is) particularly horrid. In this article in Salon, entitled “Betsy DeVos and the politics of fear: A not-so-fond farewell to Trump’s education secretary,” Adam Laats describes how

…DeVos did not care about her department because she thought her department should not exist. Like conservatives ever since the Reagan administration, DeVos yearned to dismantle her own department from within.

And:

It wasn’t just that she didn’t know anything about schools, even in her home state of Michigan. She didn’t even seem willing to learn, and she didn’t improve with time. When the COVID pandemic threw American schools into utter confusion, DeVos shocked observers once again with her trademark combination of incompetence, ignorance and chilling apathy. When asked what her department would do to help coordinate an educational response to the pandemic, DeVos punted. It was not the job of her department, she said, to “collect and compile that research.” Of course, that was precisely her job.

Laats presents an interesting abbreviated history of conservative attacks on public schools. His discussion of the battles over textbooks is particularly worth reading. He makes a strong case that the conservative case against public schools simply has nothing to do with most people’s experience with the schools, and doesn’t even involve any significant reforms. it is entirely a kind of anti-public relations campaign designed to undermine the concept of public schools as an institution.

The odd tenure of Betsy DeVos doesn’t make much sense in traditional terms. She was a department leader who despised her department, a spokesperson for public education who didn’t have any idea what to say. In more normal political times, it would have been impossible for her to keep her job. However, in the poisonous atmosphere of Trump’s White House, DeVos fit right in. Like her boss, she did not deal in facts and figures, in policies and plans. Instead, she drew on the long tradition of right-wing fright campaigns.

Artificial Insanity

Many high-profile Silicon Valley folks have voiced their concerns about the dangers of artificial intelligence and its looming effect on the technology job market. I have a considerably less bearish view on this subject, having read about AI and observed it and occasionally even tried to write AI-related code since the 1970s. But I do have my own concerns about AI. Among them is its tendency to create uncanny aesthetics.

This article links to a video that shows an attempt to train a particular type of AI to write songs – sort of. The particular type of AI is a “neural net,” trained on thousands of pop songs. Supposedly, after listening to these songs, the AI should have a sense of how pop songs usually work. So, the researchers then fed the AI fragments of the well-known song “All Star” by Smash Mouth, and asked it to continue generating a song, using what it had learned. The original song itself seems to have been in the corpus that was used for training.

The results… are disturbing. The algorithm can make some music that sounds mostly like the starting song fragment, for a few seconds, but then the result gets progressively stranger. The resulting music drifts sharp or flat, the lyrics cease to make any sense, and the song slides into deeply strange and unsettling rhythmic territory. Wait long enough, and each time, the result always seems to degenerate into some sort of cultish summoning chant in a language older than human civilization, maybe old Sumerian, or Enochian, or the guttural language of a Lovecraftian Elder God.

In other words, yes, we can create intelligence, of a sort. The resulting intelligence seems to understand some things about the structure of pop music, even if it can’t seem to maintain a lyric, time signature, tuning, or any verse or chorus structures for more than a bar or two. But this sort of musical intelligence’s most salient characteristic seems to be that, by our human standards, it is insane, a gibbering lunatic gnawing fragments of songs and regurgitating them into something horrible.

It sounds like it is suffering. Are we making artificial intelligences that suffer? I know that listening to thousands of pop songs would make me suffer. Is this experimentation even ethical?

It also suggests to me that while creating a form of intelligence using artificial neural nets may be relatively easy, the real trick might be incorporating feedback mechanisms that prevent the neural network from going insane. Maybe all intelligences based on neural net go insane by default.

Maybe this is the big challenge for human intelligence as well.

As a palate-cleanser, to eliminate the crawling horrors that may be stuck in your own neural net after listening to those musical atrocities, I recommend listening to this cover of a different song from the same era, by They Might Be Giants. It’s also a nice reminder that once, we were willing to be close to people and didn’t even wear masks!

That’s all for this time. I hope you are staying safe and that even if you are not OK, you are not insane, broken, frozen, or sick! Meanwhile, what’s on your mind? I’d love to hear from you.

About This Newsletter

This newsletter by Paul R. Potts is available for your use under a Creative Commons Attribution-NonCommercial 4.0 International License. If you’d like to help feed my coffee habit, you can leave me a tip via PayPal. Thanks!

Year IndexAll Years IndexWriting Archive