Studio711.com – Ben Martens

The Impact of AI

The other day at church, someone asked me if I thought AI was a fad. I was caught off guard. I could tell the person was probably not a fan of AI, and I had to be up front to play the piano in about 30 seconds. So I rattled off a quick answer that probably didn’t land and I headed off to play. It was a quick conversation, but it got me thinking about how people’s opinions of AI are the result of them reacting to a very narrow slice of what they’ve personally experienced.

If I had it to do over again with more time to respond, I would have started by asking what they meant by “AI.” Everyone means something different when they say it and even when they define it, it’s still easy to misunderstand them. In practice, many people mean “the free chatbot I tried.” If that’s your only exposure, I completely understand why you might decide that AI isn’t very impressive.

The free versions of Copilot and ChatGPT might not amaze you, but for coders, it’s incredible. Programming languages are a perfect playground for LLMs because the language is perfectly defined and there are countless examples for it to train on. But for other professions where AI isn’t as good (yet), it can be harder to see the value.

Anthropic (the company behind some of the most impressive AI models) recently published a paper on the labor impacts of AI.

Note that the red area is the impact that is already happening and the blue is the theoretical impact. I don’t know how they judge the theoretical upper bound, but the chart is a helpful way to frame the conversation. Will AI impact all areas of life? Yes. Will it have the same level of impact on all areas? Not even close.

The other big takeaway that is hard to grasp is that we ain’t seen nothing yet. So many areas are going to see enormous growth. For example, someone DNA sequenced their dog, developed a custom cancer vaccine, and cured their dog. It will take a while for this kind of thing to translate to humans, but I think we’ll look back on the next decade and see a noticeable increase in expected lifespan. It’s so much easier to zero in on plausible new drugs for trials when we can feed an agent tons of existing research, all the existing legislation around what can be done, and a series of desired outcomes. We’ve seen similar paths to this at work where we are researching new datacenter technology for physical tasks like improved heat dissipation.

In my area of computer programming, the change has been indescribably large. I’m blessed to have unlimited access to the best models on the planet regardless of cost, and my work life is completely different from what it was even six months ago. I have barely typed a line of code in the last three months, but I have produced more code and more business value than I ever have before. Is it perfect? No, but it is clear that people who know how to use the tools can produce a lot more value than those who don’t.

Areas that rely heavily on the generation of documents are not nearly as far along in their journey, and, according to this analysis, will not see quite the theoretical maximum impact as the more technical areas. This aligns with my personal experience as well. AI is better at summarizing documents and editing them than at creating them. I don’t write documents with AI, but I regularly use AI to review my work, suggest where the document might be confusing for readers, etc.

And finally, areas that are heavily based on physical work have the least impact now and the least theoretical impact. If I’m mowing the lawn, an LLM isn’t going to help me, but it can be valuable when I’m trying to figure out how to repair a broken mower.

As with almost all hot issues, there is more to this story than “AI is dumb. Look at this ridiculous output!” and “AI is going to solve all problems.” AI for software developers hit an inflection point at the end of November last year when two key models were released (Opus 4.5 and GPT 5.1). My work life changed almost overnight. That hasn’t happened in a lot of other areas yet, and a lot of people are judging “AI” by what they can get for free and applying it to areas where it’s not a great fit. That doesn’t make it a fad. It just means AI isn’t growing at the same pace in every field, and it isn’t equally useful in every industry.

benmartens April 2, 2026 Geek No Comments

Circuit Python Temperature Sensors

I love collecting data. Temperature is especially interesting to me and my Ecobee thermostat has three satellite temperature sensors spread around the house. I collect all that data in a database, but I’ve always been curious about having a couple more. Sure, I could just buy the extra sensors when they are on sale and have them automatically connect into the infrastructure I already have, but wouldn’t it be so much more work to design one from scratch? Let’s do it!

This is conceptually a v2 of the pool sensor that I built for Dad and Mom last year. While not without its issues, that sensor is now in a state where it works pretty well. It’s good for monitoring water temperature but I wanted something simpler for measuring air temperature and I didn’t want to do any soldering.

Those search parameters led me to the Circuit Python devices (specifically the QT Py ESP32-S2) from Adafruit and SHT40 temperature sensors. Those two easily connect with a Stemma QT cable. So that puts the total price of each device at around $20 plus tax/shipping. That’s much more expensive than this project needs to be, but I was shooting for assembly simplicity.

The code was largely written by GitHub Copilot. It’s very straightforward but the code ended up being fancier since I didn’t have to write it all by hand. Specifically the color changing LED is very informative to help give feedback about what is going on with the device.

Starting with a published 3D model, I modified it heavily to design my own custom 3D printed case for it and voila! I had a working (mostly) device! The image below is with the front cover off so you can see how everything press fits perfectly and there is a cutout for the USB adapter which provides power.

Problems I’ve had so far:

  • Connecting to WiFi was very unreliable during development. I think the tiny WiFi antennas on these devices are very sensitive to having too much or too little power. I added a config setting for adjusting the power levels that helped a bit, but I also built one that had an external antenna. This raised the cost but seems to work more reliably than my other versions.
  • I’m currently battling some mystery errors. Sometimes I’ll walk by and the light will just be solid red which indicates an unknown error. I’ve had trouble figuring out what this is, but I suspect it’s related to the internet connection. I recently added some code to reboot the devices after being in this state for a while. I’ve mitigated this by updating my data collection app to alert me if there haven’t been any new readings in the last 6 hours. When I get that, I just go power cycle the devices. Not ideal but at least I know when it happens.
  • There was a significant learning curve in flashing the firmware and switching between Python mode and Arduino mode. I’ve got it all documented now but there was a lot of frustration involved with that part. One great thing about Python mode is that the device is recognized by Windows as a filesystem over USB so it’s simple to update the code and config files. The downside is that the code itself can’t log to the file system when it is in this mode which makes debugging much harder.

So in the end I spent hours of my life and more money than if I had just bought more Ecobee sensors. But I learned something. Yay?

benmartens March 26, 2026 Gadgets No Comments

Store Brands

This is a bit of a weird post (which is saying something on this site) but as I go through the grocery store each week, I’m always looking at the prices for the store brand versus the name brands. I will almost always default to the store brand unless the name brand happens to be on sale for cheaper, but there are some items that I will always buy the name brand regardless of price. That usually gets me wondering if others feel the same way about these specific items or if there are others I should be considering. Reflecting on this is harder than doing it in store, but here are some that immediately come to mind:

  • Kleenex – Raw nose. Ouch. Enough said.
  • Paper towels – I want the smaller sheets and I don’t want them to fall apart!
  • Toilet paper – Raw… you get it. Ouch. The exception to this is Costco. Their generic brand of toilet paper works well and we had a plumber tell us that it’s one of the best ones for not clogging your pipes too.
  • Trash bags – The cheapo ones fall apart too easily. This is another Costco exception. Their generic brand of bags is great.
  • Bacon – Pay for the good bacon. It’s worth it. Locally I like to get Hempler’s center cut peppered bacon.
  • Cream cheese – This is the newest addition. I have always bought the store brand and just assumed that when a recipe says “softened cream cheese” that was just impossible to do well. Then I bought the good stuff and learned that the consistency of good cream cheese is so much easier to work with!

What did I miss on this list?

benmartens March 19, 2026 Commentary

The AI Age of Discovery

A couple weeks ago I wrote about how difficult it is to explain the changes that we’re seeing in the software development world. Any skeptics that remain, at least at major tech companies, are at risk of being a lost cause. In my own org, I’ve purposely adjusted my approach from “get everyone to try it once” to “how do we let great things bubble to the surface”. But outside the tech world, it’s nearly impossible to explain. There’s a great quote from a strong AI skeptic who recently “converted” (original language included):

The real annoying thing about Opus 4.6/Codex 5.3 is that it’s impossible to publicly say “Opus 4.5 (and the models that came after it) are an order of magnitude better than coding LLMs released just months before it” without sounding like an AI hype booster clickbaiting, but it’s the counterintuitive truth to my personal frustration. I have been trying to break this damn model by giving it complex tasks that would take me months to do by myself despite my coding pedigree but Opus and Codex keep doing them correctly. (source)

I finish work and immediately want to start working on whatever new idea I had for a project at home. I’ve literally had to forcibly put my computer to sleep so that I stop and go to bed. It’s an incredible amount of fun to be able to go from idea to working code in one evening. There is a whole world of ideas that were previously too expensive to try that are now easy experiments. As an example, in the last 10 days, here are apps that I have built:

  • Copilot Chat Export – VSCode extension that renders copilot chats as HTML for easy sharing
  • CommuteTracker – This app runs on my phone and automatically knows when I leave home and when I get to work or vice versa. It also automatically logs whether I took backroads or the interstate. Notably this was the first app I’ve ever developed in Kotlin.
  • RouteWatcher – This desktop app uses Azure Maps to determine how long it will take to get to work (or home from work) and it does this every 15 minutes with the results getting logged to SQL.
  • MlcSports – This phone app is basically a reworking of the MLC Athletics webpage, quickly showing me news for all the teams along with upcoming games.
  • TraktLite – I didn’t fully start this one within the last 10 days but it’s by far the phone app I’ve spent the most time refining. This is an alternative to the official Trakt.tv application for knowing which shows and movies I have watched or want to watch. I started with a specific scenario in mind and I’ve slowly expanded it to add more features that are tailored specifically for me and I’ve spent time refactoring the code to keep it clean.
  • WelsCallStats – Scans all of the call reports for WELS pastors, teachers, and staff ministers to generate statistics about the average call duration for each position, the percentage of calls that are accepted, the churches who have made the most calls, etc. (No I’m not publishing the stats. It’s for personal curiosity. Jon Hein and his team would publish them if they wanted to.)
  • WelsFamilyDevotions – I use devotions from the WELS at night with Elijah, but I don’t 100% love the mobile browser experience and it’s sometimes hard to remember which ones we’ve done before. This app just shows me the family devotions in order and hides any that I have read before. It also has a very clean view of only the devotion text without anything else.
  • Temperature Probe – Again, this wasn’t fully developed in the last 10 days but I spent a lot of time tweaking this embedded Python project that runs on a little QT Py ESP32-S2 board to record the temperature and humidity periodically.
  • Teams X Expander – This was a side project at work where I wanted to have a flow that would watch all Teams messages in a particular chat and anytime someone posted a link to a post on x.com, it would get the content of the post and share it in the chat so we didn’t all have to click the link to read it. It sort of worked but ultimately it was too hard to make it work within the security limitations of work apps plus the x.com API access is very expensive.
  • OneNote addon – This was another side project at work where I was trying to give GitHub Copilot access to search around all my OneNote notebooks. This has been a challenge to get working within our corporate environment and this one ended up failing too, but it was a fun experiment and I learned a lot about how OneNote add-ins are structured.
  • Interview question generator – This was another project at work that came from thinking about how to conduct interviews in this new agentic engineering world. It’s a bit silly to give people coding questions to answer, but how do I evaluate them? I used GitHub Copilot to generate easy, medium, and hard questions in three popular languages that would test how well the candidate could review code and find bugs. I was thrilled with the way this came out and shared it broadly in the company. There is a lot of discussion about how to handle interviews and I think this is a strong step forward.
  • I tried to make a tool that would convert drawings from the old “Microsoft Expression” software package into SVG. It churned on my request for a long time and eventually told me that the file format was completely proprietary, but it also discovered a way to install and old copy of it and export to something that could then be converted to SVG.

Ok, now look at that list and remind yourself that is 1.5 weeks of mostly spare time. It’s a couple hours each night. Now imagine how much I’m able to get done in my full work day on all the projects I actually get paid for! Now imagine this multiplied by 80,000 other devs (or whatever the acutal number is) at my company. And remember what I said before about new capabilities coming out almost daily that lets us run faster and do more things in parallel with less oversight.

I have always thought how cool it was to be around for the mainstream birth of the internet. I was the perfect age to start coding HTML in notepad. It was a whole new frontier and we were (in parallel with others) discovering amazing new techniques and ways to combine technologies to make cool experiences. This has been a similar feeling except now I’m getting paid to do it and the changes that took months before are happening daily now. It is awesome to get paid to learn this, make discoveries, and share them with others!

benmartens March 2, 2026 Geek

Best of YouTube

It’s time for another “Best of YouTube” post!

The first video I have for you is from an amazing musician named Jean Batiste. This was the first of his videos that I watched and it sent me down a rabbit hole of exploring his music and even his Netflix documentary.

The Veritasium channel has a lot of great science content, but I especially enjoyed this video about the physics of throwing a football.

And finally, I have 5 years of videos for you to watch from the channel Ghost Town Living. In 2020, this guy bought an old abandoned mining town and he has been exploring the mines and bringing it back to life. I started with the first video and I’ve been watching them all in order, but if that’s too much for you, start with this recap of his first year and see if it hooks you:

benmartens February 19, 2026 Links

Two Days Behind

I’m writing this as I’m still processing this excellent article: Something Big Is Happening — matt shumer. It is long, but honestly I would rather have you read that than this post.

Even for people in the tech industry, it’s difficult to explain how fast AI is improving. At work, one of my main responsibilities is literally to figure out what new tools and capabilities we can apply to our team and then help the team grow. Even with 100% of my effort focused on this, I feel like I’m holding on by my fingernails. It’s not fear that robots are taking over but a realization that things are changing faster than any of us expected. We are watching chapters worth of history books fly by every day.

For example, last week I was out sick for a couple days. The morning when I finally felt well enough to check messages, a non‑technical friend asked me what it was like working with AI. I joked that I’d been gone for two days so I was probably already behind. Then I logged in and… sure enough, a brand‑new, ground‑breaking model (Claude Opus 4.6) had dropped, and my programming tool (VS Code) had released features that make it even easier to work with multiple coding agents at the same time. I spent the entire afternoon just absorbing what had changed.

The pace of change is difficult to describe. Last summer I was mildly interested but it was clearly just a toy and most of the demos were hype. In the September, Claude Sonnet 4.5 came out and I could see how it was on the verge of being legit. On Nov 24, Anthropic released Claude Opus 4.5 and it was the inflection point. It was clear to anyone using it that there was no turning back. Opus 4.6 came on Feb 5 and OpenAI’s Codex models are surging too. People ask me what this is going to look like in a year. Who knows? I can’t even tell you what next WEEK will look like.

So yes, if you’re in software engineering, this is life‑altering in a way we’ve never seen before. But the key point is that this will change your life too. Whatever your job is, AI is already working to make parts of it obsolete. It’s a general‑purpose skill amplifier. That means whatever you’re already good at, AI can make you dramatically better and faster at it. This rewriting of reality matters for everyone, not just for people in tech. Here’s how to position yourself:

  • There will always be people around you who think this is all hype and the fad will pass. Do your best to bring them along, but the most important thing is to make sure your future isn’t tied to their denial. If it’s your management chain, find a new job. If it’s someone you’re thinking about hiring, keep looking. Denying AI’s usefulness today is like believing in a flat earth. It is provably better right now. This isn’t up for debate. Don’t waste energy arguing with people who refuse to see it. You gave them a chance to come along. If they resist, they’ll get left behind. Honestly, it might already be too late for them to catch up.
  • You might not be able to predict exactly how this will change your job, but you can keep yourself relevant by leading the way. Be the person who keeps up, uses these tools to undeniable effect, and teaches others how to do the same.

Back in December, I would still try to soft pedal all of this when I was in a group that I knew was mixed on AI. I did not want to sound dramatic or turn them off even more with my enthusiasm. But week by week, that is getting harder to do. The gap between people who use these tools and people who do not is widening so fast that it feels strange to pretend nothing is happening. I would normally end this by saying the future is now, but honestly it feels more like the future was last week and we are all just trying to catch up to it.

benmartens February 11, 2026 Geek

Painting

When we bought this house, there were a LOT of yellow rooms. We slowly got rid of them all and eventually, the only one left was the formal dining room which we let Elijah use for his Legos. It was time to paint that and we knocked it out in about a day. Two big wins on this project were having Elijah’s help and also finally having a good experience with caulking. I successfully caulked all the baseboards in the room and it went so well that I did a bunch more around the house. If you’ve ever seen me use caulk before, that will thoroughly amaze you, but I give all the credit to this video.

I still want to replace the chandelier but here are the before and after photos. The change isn’t enormously obvious in these photos, but I assure you that it’s nice to now have yellow walls around anymore.

benmartens February 9, 2026 Home Improvement

Church AV Equipment List

I’ve written previously about the 2022 upgrade of our church’s AV booth, but it feels like it’s time for another update, and I also want to include a full list of all our gear in case its helpful to anyone else.

  • Mackie ProFXV3 16 audio mixer
    • We basically got this model because I counted all our possible inputs and then wanted a little headroom. We don’t really use any of the extra functionality over the smaller versions.
  • Two PTZ Optics Move SE 12x cameras
  • Blackmagic ATEM Mini Pro video switcher
    • This has dramatically simplified our video situation. It takes care of various resolutions and framerates, records to a USB key (or connects to the computer), and allows me to perfectly syn the audio input with the video input.
    • I also enjoy having this be its own device. Some day we’ll probably move to something like OBS running on the PC, but for now, it’s really nice to have a device that “just works” without worrying about any updates or configurations.
    • This device also supports streaming directly to YouTube or Facebook but we haven’t started that yet.
  • Shure microphone setup
    • Pastor:
      • BLX1 Wireless Bodypack Transmitter
      • Shure DH5 DuraPlex Omnidirectional Headset Microphone
        • This is an upgraded microphone that fits Pastor better.
    • Shure BLX24R/SM58 Wireless System with Rackmountable Receiver and SM58 Microphone Capsule Band H10
      • We usually have this by the piano. We aren’t often recording/posting piano stuff but it helps us amplify the piano volume during the services to make sure people can hear it while they are singing.
  • Mackie C200 speakers
  • QSC GX3 300-Watt amplifier
    • I would not buy this again. There’s nothing wrong with the product, but it’s extremely overpowered for our space. We usually leave it on volume 2/10 and I’m pretty sure that if I turned it up much past 7 I could damage the building. We were coming from an extremely underpowered system so I guess I overshot the goal.

So our whole setup is PC independent. Once the service is done, we take the thumb drive out of the video switcher, plug it into the PC, and upload to YouTube.

We also updated our internet service from DSL to business cable and now our service is reliably at least 200Mbps symmetric. Uploading is usually done before we can finish entering the metadata!

We have a TV in front of church that is mostly used to show quick videos after church but is rarely also used during a sermon. We can control that from the balcony using an HDMI over Ethernet extender. This adds about 200 ms of video lag but we take care of this by using VLC for playback and adjusting the synchronization in the VLC settings. The TV in front of church is set up as a second monitor on the balcony PC so we can prep content privately and then drag it over to the TV in front of church. Via an HDMI splitter, that front TV feed also gets sent to an input on the video switcher. This lets us show the sermon PowerPoints directly in the video recording with perfect quality.

We have a great crew that is trained to operate this booth. Our sweet spot is around 4-6 volunteers. That is big enough that we don’t feel like we’re doing it too often, but we also do it often enough to remember how it works.

If you’re curious to see what it all looks like, check out CalvaryLutheranWA – YouTube. I edit the shorts after the fact, but pretty much everything else posted on that channel comes straight out of the AV booth with no post processing.

If someone gave us a bunch of money to upgrade this, I could easily spend it, but unless we get asked to start live streaming the whole service or something like that, I think we’re in a pretty good spot with our setup … for now.

    benmartens February 5, 2026 Geek

    Homemade Air Filtration

    Many years ago, I duct taped an air filter to a box fan and hung it from the ceiling above my workshop area in the garage. It’s cheap, but it was so effective that I ended up having another one that I move around the shop when I’m doing a particularly dusty task. (Note that it is a bit of a fire hazard to purposely draw some dust right into your fan motor so use at your own risk. The filter doesn’t catch it all.)

    That has worked great for years but it’s a pain when I change the filter because it’s hard to get the old tape off and then the new tape doesn’t want to stick to the dusty fan. After thinking about ways that I could model and 3D print some kind of clip, I realized that others had surely solved the problem before. Sure enough, I found a great model online that included the ability to easily resize the clip to fit my specific fan and filter. After a little trial and error, I got it sized just right and the clips are awesome! This is so much better than the old tape approach. I did have one of the clips break when my mobile version of this fell over so I might reprint these to be a bit thicker.

    benmartens February 2, 2026 Maker, Woodworking

    Doom Scrolling Thoughts

    I recognize that this is a loosely supported hot take, but you know how we look back at smoking in the early/mid 1900s and wonder how they could have done that to themselves? In 30 years, I think we might look back at social media like that.

    I’ve been chewing on this for a while and it’s one of the main reasons why I stopped using social media in general but more specifically, I avoid any app with an infinite scroll. That mind-numbing flick, flick, flick feels like the exact opposite of what we’re told about how to decrease the risk of cognitive decline (Alzheimer’s, dementia, etc.). Much of the research into those diseases is inconclusive at this point and then I realize it’s another big leap to say that doom scrolling makes it worse. But it seems illogical that repeatedly losing track of time while scrolling through 10 second videos is going to strengthen your cognitive abilities. Smoking one cigarette doesn’t give you lung cancer and flicking through videos on the toilet one time isn’t going to rot your brain, but what does prolonged and repeated infinite scrolling do to you?

    This is going to take the scientific community a long time to study, but there are already papers available that link doomscrolling to poor memory, shorter attention spans, impaired decision making, and cognitive overload. In kids, it has been linked to increased rates of anxiety and depression. Links are one thing, but is it correlative or causal? This is harder to figure out but there is evidence for doom scrolling being causal. There are some studies that survey how people feel and perform after being exposed as well as fMRI studies showing craving reactions and abnormal prefrontal cortex activity. There are also strong causal links between doom scrolling and increased stress (cortisol) levels which has major known negative effects on long term health. At the same time, technology can massively improve your mental health when used correctly. There is a difference between passive scrolling and active, intentional use.

    “All in moderation” is a motto I support, but after setting up an app on my phone that tracked my screen time, I realized that I’m not able to do infinite scrolling in moderation. I’d rather err on not using it at all.

    More reading:

    benmartens January 29, 2026 Uncategorized