Nels Lindahl — Functional Journal

A weblog created by Dr. Nels Lindahl featuring writings and thoughts…

Month: March 2023

  • Consider what research trajectories to really deep dive

    Yesterday I started to consider what research trajectories to really deep dive into this year. The next few weeks are going to be devoted to some literature reviews within the polling methodology space. That is an area where I want to dig in and see some contemporary research. I have some quality content written for the next couple of Substack posts, but that writing needs some time and attention to get it over the finish line. It’s the continued release of all these new models that keeps pulling my attention in other directions. Pacing these releases to any normal schedule would be impossible at this point. So many companies are trying to put forward chat enabled bots that it is nearly overwhelming. Right now I have been using the ChatGPT version from OpenAI and Google’s Bard. I’ll admit right now that the one that I interacted with over a longer duration was the OpenAI version. It has helped me write a couple of books that are passable. One on ethics in AI and the other about political debt. The one on AI ethics will show up in my Substack post today. 

    One of the things that I have considered a bit is how I’m picking titles for these weblog posts. Given that the only social media they are shared on is Twitter as a single tweet. Maybe I should be crafting the title of the posts as a stand alone tweet. Right now a title is simply selected from some interesting block of words (more than 3) that were written as a part of the post. Over the long history of this blog a lot of good titles have been used and now we are at a point where generating unique ones is a bit more challenging. To that end it has become easier to just select from text a passable title and move along with the writing project. Recently, I have been highlighting where the title came from in bold text to help make that title selection reality a little bit more obvious to the reader. 

    Tomorrow morning is going to be key in terms of recording and editing down a couple of Substack posts to get back on track in terms of a backlog. I had successfully worked ahead and now I’m back into a position where all the content that is complete will be published. Later today the last of the recorded and ready Substack editions will go out for publication. I have been working on the next ones of course and that content has had review and effort put into it, but none of it has been recorded. I have not had my microphone setup for the last two weekends making it harder to record high quality audio. I had considered trying to use my Pixel 7 Pro smartphone to do the recording, but it does not sound as good as the Yeti X microphone. A huge difference exists in the audio quality on those two microphone systems. One is markedly bigger than the other. The Yeti X is bigger than my phone to begin with and the microphone on a smartphone is exceedingly small as a component.

  • My normal routine got a little bit out of rhythm

    Things within my normal routine got a little bit out of rhythm. For several days I turned off my alarm clock and just slept until either something woke me up or the day started off naturally. It has been a long time since that type of start to the day had even been considered. Even after returning to a normal sleeping schedule it took a couple of days to get back into the mode of starting the day with a little bit of writing. Yesterday it snowed at the end of the day which was surprising. The snow did not help me get back into the habit of daily writing, but it apparently did not hurt that pattern of effort either. 

    This month has been truly bonkers in terms of language models being released by major players in the AI space. Things have been so active that I’m actually considering changing up my writing plan for the next few weeks to provide a bit of coverage. That change up probably won’t be something that happens, but it is interesting to consider. I may add some of these topics to the very end of my backlog to reflect on in a few months. Providing immediate reaction coverage is not really what I need to do with my time. One of the things that I find the most challenging about our news feed based communication system is that it values recency over accuracy. We have access to a steady and perpetual stream of news, but few things really get covered over a long period of time. 

    It is entirely possible that these new models being introduced right now will be impactful enough within the day to day lives of people working, making things, and otherwise engaging digitally that it has to become a part of things. Even Bill Gates jumped into the argument and noted that this current wave of technology is the most important change in decades. It is a lot to consider and my writing session here is drawing to a close. It’s about time for me to jump back into the curation of my topic background and give the road forward a deep bit of consideration. I will note here that it would just take an hour or so for ChatGPT to produce the content for my entire 2023 writing backlog. I like to think that my efforts would be superious and that ChatGPT could not match the depth of my reasoning and consideration, but some of the content being produced is passable. It reminds me not to produce middling prose and to really value the creation of deeper works of content.

    Maybe it would be a good time to shift my writing focus into a pure deep dive of polling methods, civic engagement, civility, civil society, and the social fabric of things that bring us together. I’m looking at my backlog and thinking that I should probably just go deep on some topics and blow out my backlog topics toward the end of the year. It would be pretty easy to just refocus a bit away from purely writing about AI and begging to evaluate AI and something else within the focus related to society in general. That effort will be easy enough to implement and we will see what happens here this weekend. I’m about 80% done with a couple posts that need to be finished up and recorded. Working those projects to a conclusion will let me set up the next phases of this writing experiment.

  • Thinking about good adventure books

    Instead of working on a Substack post right now I’m just listening to the WAN show and drinking a glass of water. That is how this day is starting off. I spent a few minutes looking up all time great adventure books to find a good one. Most of the time I read non-fiction books so I have not really gotten that far into many other books. Recently, I read the entire The Expanse series which is the first modern series I have consumed outside of the derivative Dune books beyond the original series. So many books are getting published right now that it is hard to know what to read next. So many people read that Ready Player One book a few years back that I picked it up and thought it was great. A lot of people recommended The Expanse series so I gave it a shot and it turned out to be delightful in terms of being a space opera with reasonable technology. Obviously, I would have preferred that many novels be devoted to extending the Firefly universe, but you cannot always get the things you want. 

    To just be clear about my point here, I’m thinking about good adventure books. Trying to figure out what audiobook to read next is a bit of a time consuming thing. To be clear I don’t mind paying the audiobook price where the royalty goes to the publisher and author. They did some work and I plan to enjoy listening to it or sometimes reading it. I’m just not in favor of buying audiobooks that are disappointing. Most of the time I’ll be listening to a podcast and they will recommend some books to read. I’ll open up Google Books and buy the audiobook to listen to in that hour before bed. The vast majority of those books are basically modern history. Finding really good science fiction series has been a bit harder.

    I’m open to recommendations for a good read. Generally, I prefer epic space journeys.

  • A golden window of writing adventures

    Things got off to a better start today. My energy level was higher at the start of the day. Rocky dog did kick off that day a few minutes before my alarm clock should have gone off. Apparently, the dog based alarm clock was running a little bit early. It’s entirely possible that next week will usher in a golden window of writing adventures. We will see how that one goes. I should have some really good blocks of time to engage in some active writing. That does not always equate to a higher degree of productivity. It just opens the window to it. That opportunity has to be realized to materialize anything. One of the things I want to spend some time working on next week is deeply considering my 5 year writing plan to evaluate where things are and what things should be changed up. 

    Throughout this month I have been pretty consistent about sitting down and producing some blog related content. Maybe a bit more of an ongoing narrative would be helpful. Perhaps at the crux of this problem is that each one of these missives is the opening part of getting started. Functionally it is a similar thing that happens every day as the process of going from being awake to being ready to be productive starts. That is what I’m attempting to rationalize anyway. An ongoing narrative would be related to the things that I’m interacting with and an explanation of things that are happening along the way. Probably the only way to really do that would be to make sure that a bunch of starter ideas are seeded in Google Keep throughout the day. At the start of the day instead of writing without a prompt. I really just open a blank page and begin to make my way into the writing process. Things would instead start out by grabbing the previous day’s seeds and beginning to work bringing form to those ideas. 

    That method of seeding would certainly produce an ongoing narrative that was essentially about my interactions with things. It would push this from being a more functional journal and into a more reactionary journal. For better or worse that might be a more interesting thing to consider for a couple of weeks. I pretty much did that during the million word challenge where I had a backlog and used that to help focus my writing efforts every morning. During the course of trying to produce 3,000 or more words per day for a blog that is going to be required. 

  • I just sit down and make words appear

    Substack posts are queued up until Friday, March 24, 2023. That means the next 3 Friday nights will receive a post. This morning as I sat down here to begin the day pretty much everything except something productive got done. I was checking emails, looking at websites, and generally just not engaging in the act of writing. It took about 19 minutes of that brand of nonsense to complete before this Google Doc was opened full screen and the writing process took hold. I tried to use both Microsoft Word the desktop edition and the online version for my daily writing exercises. It’s really one of those things where I strongly prefer to write in the Google Doc’s interface. All of my editing and manuscript preparation happens over in Microsoft Word for books and articles that are not LaTeX based. Anything that falls into the LaTeX required camp of writing is being formatted in Overleaf online. It’s not being written in that editor, but the formatting happens post content creation for publication using that platform. 

    The interesting part about the whole Overleaf thing is that I pay them every month. The rate just went from $5.00 to $5.50 to have an online collaboration friendly LaTeX editor. Intellectually I know that I could install a LaTeX editor on this Windows desktop and not pay that subscription fee, but in practice it seems like that is unlikely to happen. I move around between Chromebooks, Ubuntu, and Windows depending on where I’m working at the time. That makes having the LaTeX related content accessible from a browser pretty darn helpful. We will see what happens if they keep raising the prices. We may find out at what price I’m not willing to keep paying them for their services. I’m sure the body of academics they have using the platform are mostly using the sharing functionalities that Overleaf has created. If you were working on a paper with a collaborator or several, then it would be very convenient to share it this way. Learning how to be a copy editor for my own work using LaTeX was an interesting experience. It’s a steep initial curve and then you sort of get used to it. 

    Writing in a browser like this is a much easier experience. I just sit down and make words appear on the screen with the keyboard while moving along with my day. Having to sit back and think about commands and formatting while that happens breaks my concentration and is disruptive. That is probably why I edit and create my content in a normal word processing document. I have been doing much better in the last 30 days in terms of focusing on creating some words for the blog at the start of the day. Part of my routine is to focus my energy for a few minutes and collect my thoughts at the start of the day. It really is a significant part of starting the day off right. Being focused at the start of the day helps me use my most productive time and golden hours on the right things instead of the easy things. That might seem trivial, but it is really pivotal to getting things done. Productivity for me during those few hours where no interruptions occur and work can progress unabated is higher and perhaps a deeper kind of work. 

    We are at a time where things are changing very rapidly. That change may end up being truly disruptive to civil society and the way we interact with each other. We have already seen a tremendous change to the way news is consumed, created, and shared. A study of information exchange and civil society would be an interesting way to spend some time. That is probably just adjacent to my research interests, but could perhaps be a part of some of my upcoming research. We will see what happens on that front as we move forward. It’s certainly something that is front of mind for me and could get included at some point. 

  • Cleaning up the blog a little bit at a time

    Things did not start off very quickly today. Perhaps that is an understatement. Things started off painfully slow today. I’m not entirely sure why that happened. The dogs were a little restless last night. My sleep routine is pretty good, but disruptions do seem to create havoc on my day. One of the interesting features of the Oura ring I acquired the other day is that it tracks my sleep patterns pretty closely. I’m enjoying the ring so far and consider it to be a solid purchase. We will see if that type of recommendation holds up after a year. 

    Even writing that last paragraph took a lot longer than it should have to type out those words. Maybe it is time for my efforts at stream of consciousness related prose to give way and consider something more directly. Over the last few days I have been cleaning up the blog a little bit at a time. It’s requiring a bit more effort than I expected, but so far it seems to be worth doing. One of the big things that I need to do is work to remove the “Upcoming Research” page and combine that content into the “Research Trajectory” page. The pages on a blog are stand-alone static content repositories and are not a part of categories or archives in the same way a regular post would be. They are not date driven items they more or less stand on their own instead of being part of a collection. That being said, I don’t need both an upcoming work page and a trajectory page. It’s really the same thing said in two different ways. 

    I went ahead and moved the page to the trash section of the blog. It’s no longer publicly visible and will be deleted shortly. It contained the following:

    Upcoming Research

    This is just a list of upcoming research paper topics that I have started to sketch out. 

    1. Open source MLOps paper (from talks)
    2. eGov 50 revisited paper
    3. Local government technology budget study
    4. The fall of public space paper (could be a book)
    5. A paper on the quadrants of doing
    6. A brief look at my perspective on interns
    7. Some time of perspective on the audience size of ML and why…
    8. ML model stacking
    9. Something on reverse federation
    10. A hyperbolic look at the conjoined triangles of ML
    11. A literature review of modern polling methodology
    12. A literature study of mail vs. non-mail polling methodology in practice and study
    13. ML mesh
    14. A paper on political debt as a concept vs. technical debt

    It looks like all of those topics have been moved into the research trajectory list. Strangely enough the last 5 year writing plan that was loaded up was from March 3, 2022. I know that the content got reworked during week 104 of The Lindahl Letter and I’m going to load the more recent version. Ok, that update had been completed. It looks like my top 5 research interests need to be updated, but outside of that the page is now better updated. 

  • I have been sending out tweets for a really long time

    Yesterday was one of those days that I was super duper disappointed in Twitter. Generally I don’t deal with a whole lot of trolling on the internet. Maybe that is just the long tail of statistical averages catching up with me. That is entirely possible. Maybe somebody was just having a bad day and I ended up having the tweet that caught all that emotion. It did not really emotionally impact me as much as it did surprise me that so many other people engage in the drama. Maybe that is the essence of Twitter and I just don’t see it very often. Most of my engagement is with technology, machine learning, or academic Twitter communities. Not only was it a strange day on Twitter, but it was also my 14th year anniversary on the platform. Apparently, I have been sending out tweets for a really long time.

    For the most part tweets are functionally ephemeral; they occur and then fade away from attention. Sure somebody could really dig to go and find an old one, but practically they just fade away from the feed and that is that. I’m not entirely sure that I get anything from the 10-20 daily minutes I spend on Twitter. Maybe that time should be noted as entertainment time and it should be managed as such. My take on Twitter overall is that it happens to be an ephemeral platform. People will crowd to it as a source of first news given the overall reach and opportunity for people to actively report news. The barrier to entry on being able to report news is very low. Given the current state of local news coverage nationwide it is not entirely surprising that crowdsourcing news is a staple of modern coverage. That aspect of things will probably keep crowds of people flocking to the application to tweet and check tweets. To that end, unless some other platform shows up for people to report and share current events I imagine that Twitter will continue on in this current form for a while. 

    It might be easier to just fip my Twitter profile over to private mode, but that would diminish engagement. Most of the time my engagements with others are positive. Overall I have been running my phone in the do not disturb mode for months now. People who are on my contacts list can call and break into my day. Otherwise it functionally shuts down all notifications, sounds, pings, or other indicators that some application wants my attention. In terms of getting deep work done and being more productive it has worked out really well. At one point, I had even considered getting one of those new flip type folding phones to really make it harder to activate the screen and engage. I might end up going an entirely different direction and getting one of the phones that sort of folds open and becomes a small tablet. That seems like a pretty decent use case for reading things. The price on those types of devices will probably be trending downward shortly as they have been on the market for a couple technology generations. 

  • Still creating massive missives of prose

    For some ineffable reason yesterday I thought it was a good idea to move the widgets around on the blog. If you had to pick one thing that could be done, but had very little end value that would certainly be on the list. Most people who encounter the blog got to it via some sort of search for a keyphrase that I happened to write about sometime in the last couple of decades. Maybe they will use the search box to look for more or maybe they will just bounce after reading whatever it was that brought them to this platform. Based on the statistics a few of them click around from time to time trying to figure out who is still blogging after all this time would be my guess. Writers write and the ones that got into the habit of blogging are probably going to be the last ones still creating massive missives of prose

    In the absolute calm of my morning writing routine I did pull the “Obligatory Search Box” from “Footer 3” to the normal “Sidebar” position. That puts a search box at the top of every desktop viewers page and right after the opening text of whatever page somebody would land on within a mobile browsing experience. From what I can tell the mobile experience basically displays the page in an order from top to bottom assuming an order of header, body, sidebar, and footers 1 to 3. My very earnest guess is that nobody outside of me uses the content in the footers. I happen to like to watch the “Weblog Calendar” fill up as the month goes on and the “Very Random Tag Cloud” is obviously a self-serving way for me to be reminded that the range of my writing needs to be more expansive during these blog posts. 

    My writing has to shift into the category of being more active to describe the things that are happening and what is going on in terms of my interaction and reaction to them in the form of prose. A lot of what this blog ends up being is a reflective look at the process of writing and the nature of how the conflict of action and inaction gets processed both for me at the time and against my view of what writing trajectory I’m supposed to be marching down at the time. Some actual pruning did occur with 2 pages being moved to the trash bin within the blog. Both “Podcasts” and “Guitar Gear Watch List” were moved to the dustbin of history. 

    I’m sure you really want to know what was on the podcast page so here it is:

    Here are my top 5 favorite podcasts as of June 23, 2019:

    1. Tomorrow Podcast
    2. This Week in Google
    3. The Vergecast
    4. This Week in Tech
    5. This Week in Science

    I’m listening to fewer podcasts right now. Here are my top 3 favorite podcasts as of June 28, 2020:

    1. Tomorrow Podcast
    2. This Week in Google
    3. The Vergecast

    At the moment (0741 hours, Saturday, September 5, 2021), my podcast interests have changed a little bit. This list seems to change a lot… Right now I’m still using Google Podcasts as my primary listening application.

    1. The Vergecast // Decoder podcast
    2. Lex Fridman Podcast
    3. Machine Learning Street Talk
    4. Science Friday
    5. FiveThirtyEight Politics

    To round up this post to current and bring it fully into the now here are my current top weekend listening rotation podcasts as of Monday, March 6, 2023 at 0540 hours:

    1. All-In with Chamath, Jason, Sacks & Friedberg
    2. Hard Fork with Kevin Roose and Casey Newton
    3. The Vergecast & Decoder (both are Nilay productions)
    4. Machine Learning Street Talk
    5. Lex Fridman Podcast (I’m 50/50 on listening to these) 
    6. New Heights (everything reached peak Kelce in the playoffs)

    As an alternative to listening to podcasts you could listen to one of my curated Pandora internet streaming radio stations:

    1. Warren Zevon Radio — https://www.pandora.com/station/1406490537159530917
    2. In this Moment Radio — https://www.pandora.com/station/2702008915953424805
  • Content that wandered into existence

    Yesterday I ended up recording episode 113 which will debut Friday, March 24, 2023. This is the single longest podcast recording of anything in that entire series. It ended up coming in at 26 minutes of recorded content. I did the recording in one take. Yes, I was oddly proud of being able to do the entire recording in one take. While I’m writing this particular post we are sitting at a point where the next 3 weeks of posts are ready to go and will go out assuming Substack continues to exist. 

    One of the things that I did spend a few moments thinking about was if I should just be posting the content on the blog when I get done creating it. At the end of the year, I’m certainly taking the content and putting it into manuscript for publication. At this point however, I’m not posting it on both the blog and substack. They are for the most part distinctly different communities and I certainly could cross post the content. I’m not sure why during the course of finishing up the 113th week of content that wandered into existence

    During the course of my efforts this morning I looked at the next 5 weeks of content and started to sketch out how things were going to come together. Today could very well be a day where I have some time to sit and research for a few hours. We are probably going to watch an SNL replay given that a Kelce brother is hosting it.

    One of the other things to note today is that I have reduced my caffeine consumption. Today started off with one shot of espresso and a glass of water. Instead of having 2 shots of espresso the day has officially started with just one shot of espresso. This was a major 50% reduction in my daily start of the day caffeine consumption.

  • Lamenting some smartphone repairs

    Yesterday before noon the repaired Pixel 7 Pro smartphone from Google arrived. For the last couple of weeks I had moved back to my Pixel 5 which remains a pretty good smartphone. Apparently, they repaired the phone that I sent back to them via a nice cardboard package they sent me for traveling with a postal carrier. The phone I received back has the exact same IMEI as the one that was sent to them for repairs. Based on that I’m pretty sure they actually repaired the unit that was sent in for repairs and did not issue a replacement phone. I took a close look at the actual phone and didn’t really see any evidence that they heated up the glue that holds the screen in place and got inside it to make some type of repairs. They had it from February 21 to February 28 so it is entirely possible that they did take it apart and effect some type of repair to the screen. I’m not entirely sure what happened and to be honest nothing along the way shared any details about what exactly went wrong or what they did to correct the problem. 

    Based on my observations something within the screen went very wrong. When you hit the power button the top half of the screen would flash with a dark green sort of tint or the whole phone would flash white and then nothing would happen. The phone itself was probably running and waiting for some type of user input, but the screen was dark and unresponsive. A few people certainly encountered this and I tried to look around for solutions or at least other people to chat with you had experienced the same problem. For the most part, when a smartphone behaves like that people send them back for a replacement. In my case the warranty part of the service cycle would have done that for me and has been fine before. I used it when my screen got cracked. They pretty much sent a new replacement phone and I sent the cracked one back to the warranty center. This time around the warranty and repair was done by Google. 

    I loaded the data, contacts, and well everything from my Pixel 5 to the newly repaired Pixel 7 Pro using a standard USB Type-C cable. I pretty much only use official cables from Google chargers that I ordered from the Google store online. The only other charging mechanism that I happen to use is a Google stand for charging. My preferred method of charging is to set the phone on the wireless charging stand. It’s so much easier and used to feel like the future, but now it is more routine than anything else. Completing the data transfer process included moving about 62 gigabytes of data between the phones. It took around 20 minutes and then I had to sign into all the applications which took another hour of time. Getting all the authenticator stuff changed out took maybe another 15 minutes. It was a quick reminder of why I kept my backup phone in the first place. Getting things back in order without some backup would be a lot harder. Based on this exercise about 2 hours of time needs to be reserved to complete the phone transfer process. 

    This post might come across as if I’m lamenting some smartphone repairs and that truly is the case. Overall my trust in the Pixel line of products has been diminished. I was a day one device order for the first 5 generations of Pixel smartphone products. I sat out the Pixel 6 as my enthusiasm for it was limited and some of the initial reports conflicted about the build quality. I jumped back in and ordered two of the Pixel 7 Pro phones and the Pixel Bud Pro’s to go with the phones. Maybe delivering hardware is not the thing that Google is most focused on achieving. My Fitbit certainly has not gotten any great updates or support after the acquisition by Google. I actually switched to the Oura ring recently and abandoned using my Fitbit smartwatch. We will see what happens within the organization changes at Google over the next year. I would not be surprised if hardware fades away altogether.