A few thoughts on Cal Newport’s Digital Minimalism

Cal Newport: Digital Minimalism

  • I’d heard about Cal Newport when he was on Ezra Klein’s podcast about four years ago. I’d always had some sense of his argument – that we all have unhealthy relationships with our phone and technology – but hadn’t really done much about it.

  • After I listened to that podcast, I tried a version of his digital detox – deleting all social media apps from my phone, forcing myself to not use the browser. But I still used Google Maps, Kindle, and texted.

  • After that 30 days, I got my daily iPhone usage down to less than an hour a day. I’ve reverted back to a higher level, but I still go in cycles of adding/deleting Twitter and Instagram from my phone. They are both huge distractions to me, especially Twitter. The only problem is I find myself sometimes LinkedIn in its place when I have a free moment. LinkedIn!

  • I’m not worried about maintaining only weak connections through social media – because all my social connections are weak at this point in my life. I have kids, I literally have no time to hold phone call office hours or keep an hour open to grab coffee every week. Time is absolutely my scarcest resource, and it’s not like I’m already frittering it’s away on Instagram.

My biggest takeaway: I need to spend more time creating and less time consuming

My job (a macroeconomic analyst) is a lot of creating. I write a lot for my job.

But I need to do better creating outside of work.

  • I look back at some of my old blog posts from 2014 and am impressed by the breadth of what I was thinking about (for example, I still get traffic on a post about inequality mapped through broadband availability. I’m not sure how – I thought all my work then was offline. I need to look at that…). And it was all writing that was just for fun.
  • There’s no better way to understand something than writing about it. It forces you to get deeper and understand the nuances.
  • This is especially important as hot Twitter takes drive the narrative, which are the most superficial way of communication.
  • Twitter is hugely important to understanding what broad ideas are out there, but you need to do more than write/read 280 characters to really understand it. That should be obvious but I’m practice that’s how people operate.

It’s a really good book. A lot of the advice may be impractical for a lot of people, but I think it is directionally correct for most people. How we need to not only do better with our devices. And especially how the idea of the “slow news movement” would be much healthier for all of our information consumption.


November Update: Smart contracts and Solidity

I made good on my promise to myself to learn more about crypto and blockchain.

My last month was mostly filled with the Buildspace Solidity + Smart Contracts project. It was really well done and I look forward to taking more from Buildspace.

Am I going to go all in on Ethereum after that project? No. I still think that a lot of the most mentioned crypto and blockchain use cases don’t even require crypto or blockchain at all. It’s more often than not a solution in search of a problem.

But I hope the whole ecosystem continues to develop because there may be some value to come out of it.

What are my biggest takeaways after the Solana project?

First, I like to think I’m fairly technically minded. I’m not a software engineer by any stretch, but I really try to understand the ideas and concepts behind the software and code that drives technology and ideas and business forward.

But I’m not representative of everyone – and there’s no way that smart contracts, at least in their current form, can take off broadly. There’s a big UI and UX problem with blockchain.

Second, I’m trying to understand how this scales more broadly. I’m currently reading Nathaniel Popper’s book about Bitcoin, and even a lot of the early Bitcoin advocates didn’t trust themselves with private keys. And a lot of them are actually developers! So what’s the point of decentralization when peoples own preferences and tendencies drift toward centralized organizations?

What’s next?

I would really like to make an effort to finish the AWS Machine Learning certification next. I also would like to brush up on some of my Python and SQL fundamentals but I felt like I lost momentum on AWS ML with the blockchain project, and I need to focus on one thing at a time.


The decentralization movement has a messaging and UX problem

I own zero crypto or digital assets. I have no horse in the race – not because I don’t believe they’re important, or part of the future of finance, or a technology that we may use every single day at some point in the not too distant future.

I simply just buy passive broad-market ETFs. I’ve handcuffed myself to that strategy (but that’s another post for another day).

But I am trying to learn more about the crypto/NFT/digital asset space. I am finding I have a lot more questions than answers as I go through it. I’m a lot more positive on a lot of parts of the space, while I’m a lot more negative about other elements.

There’s a user-friendly design flaw across the decentralized space

I apologize for using “user-friendly.” It’s a term that has been bashed in the ground as everyone and everything tries to copy the iPhone – one of the most complex machines ever invented, yet so intuitive that it doesn’t require ANY instruction manual.

Advocates of crypto and NFTs and blockchain technology think they are sitting on something that has the same power to change the world as the iPhone. I can’t judge whether or not that’s true. But the “blockchain fixes this” meme quickly went from earnest observation to a (fantastic) Twitter joke. Not because blockchain technology can’t solve a lot of problems – it can – but because the current design flaws of the space and the messaging around it are so absurd.


I was recommended a book to read to understand the broader… let’s call it ethos of blockchain.

Amazon link

I think the book fails in a few regards. First, I don’t think it does a very good job of explaining what blockchain really is.

It does an even worse job producing some use cases for blockchain – examples which are fanciful at best and outright blind to any sort of real-world perspective at worst.

Here’s a snippet:

Digital Asset Proof as an Automated Feature In the future, digital asset protection in the form of blockchain registry could be an automatically applied standardized feature of digital asset publication. For certain classes of assets or websites, digital asset protection could be invoked at the moment of publication of any digital content. Some examples could include GitHub commits, blog posts, tweets, Instagram/Twitpic photos, and forum participations. Digital asset protection could be offered just as travel insurance is with airline ticket purchases. At account setup with Twitter, blogging sites, wikis, forums, and GitHub, the user could approve micropayments for digital asset registration (by supplying a Bitcoin wallet address). Cryptocurrency now as the embedded economic layer of the Web provides microcontent with functionality for micropayment and microIPprotection. Cryptocurrency provides the structure for this, whether microcontent is tokenized and batched into blockchain transactions or digital assets are registered themselves with their own blockchain addresses. Blockchain attestation services could also be deployed more extensively not just for IP registry, but more robustly to meet other related needs in the publishing industry, such as rights transfer and content licensing.

Ok, that’s all great! Here are a couple of real-world use cases where an immutable record would be useful and important.

But then it just starts rolling downhill from there:

There could be “personal thinking chains” as a life-logging storage and backup mechanism. The concept is “blockchain technology + in vivo personal connectome” to encode and make useful in a standardized compressed data format all of a person’s thinking. The data could be captured via intracortical recordings, consumer EEGs, brain/computer interfaces, cognitive nanorobots, and other methodologies. Thus, thinking could be instantiated in a blockchain — and really all of an individual’s subjective experience, possibly eventually consciousness, especially if it’s more precisely defined. After they’re on the blockchain, the various components could be administered and transacted — for example, in the case of a post-stroke memory restoration.


As mentioned, in the vein of life logging, there could be personal thinking blockchains to capture and safely encode all of an individual’s mental performance, emotions, and subjective experiences onto the blockchain, at minimum for backup and to pass on to one’s heirs as a historical record. Personal mindfile blockchains could be like a next generation of Fitbit or Apple’s iHealth on the iPhone 6, which now automatically captures 200+ health metrics and sends them to the cloud for data aggregation and imputation into actionable recommendations. Similarly, personal thinking blockchains could be easily and securely recorded (assuming all of the usual privacy concerns with blockchain technology are addressed) and mental performance recommendations made to individuals through services such as Siri or Amazon’s Alexa voice assistant, perhaps piped seamlessly through personal brain/computer interfaces and delivered as both conscious and unconscious suggestions.

I don’t know if it’s because this book was published in 2015, at the true peak of technolibertarianism, but I don’t see these kinds of arguments out there in the wild much anymore.

A quick pause to quickly define terms. I’ll use privacy here not as hiding one’s self from others, but removing one’s self from entrenched institutions: businesses, government, etc.

I think that privacy has become the ultimate goal of the decentralization movement. That’s fine, but as we’ve seen with countless examples in the past years (hacks, scams, grifters…), just because you operate on a decentralized platform, that certainly doesn’t mean it’s private. Are we really going to put our minds on the blockchain (ignoring the fact for the moment that that’s nowhere near possible)? Are we really going to log all of our vitals and most important personal data about ourselves and let it live digitally, even if its on the blockchain? The author admits this is science fiction, so why even mention it?

Or better yet, why would blockchain technology even be the best solution for this type of problem?

Bring it back to Steve Jobs and Dieter Rams

One of the first Mac ads said:

Since computers are so smart, wouldn’t it make sense to teach computers about people, instead of teaching people about computers?

And the industrial designer Dieter Rams wrote up 10 principles of good design.

A few key choices? Good design…

  1. makes a product useful
  2. makes a product understandable
  3. is unobtrusive
  4. involves as little design as possible

That’s it. That’s what’s missing with crypto and the broader decentralization movement.

If you want to promote blockchain technology and cryptocurrencies and start NFTing everything, you need to start producing not only reasonable use cases, but better user experiences to achieve it.

Currently cryptos and NFTs are being designed by crypto/NFT advocates for crypto/NFT advocates. There aren’t a lot of people out there putting themselves in the typical consumer’s shoes and making those experiences seamless in our day-to-day lives.

(I’m also not paying a 21% fee to buy ETH, but again, another post for another day)


October Update – Python, AWS. Next up, Blockchain?

My output on this blog has been less than ideal. I’d still like to use this as:

  1. A way to keep track of my own progress in developing new skills and ideas (and putting it into public view as a sort of motivating factor)
  2. A portfolio of some of my work

But I feel like I have a good excuse, at least this time (though I should remember to set aside time every week to regularly reflect on where I am and where I want to go).

A productive two months

It has been almost two months since my last entry. But it has been a very productive stretch. A rundown of my last 6 weeks:

Earned: Python certification

I received my company’s internal Python certification. It was built mostly around data analysis using Pandas/Numpy and building data visualizations with Seaborn.

A lot of the prep was done with LinkedIn Learning courses, which I thought were fairly high level, but fine. It definitely wasn’t as in depth as something like the Python for Data Science and Machine Learning Bootcamp course by Jose Portilla on Udemy, which I used to jump into Python for data a few years ago and thought was excellent.

This was stuff I’ve had experience with before, but good to know I’ve cleared my company’s benchmark.

Pending: API certification

I also completed all of the work to complete my company’s API certification as well – I’m just waiting on being judged.

This program was centered on building tools using our set of prebuilt APIs for clients. It was as much or more geared as a background refresher for sales staff as it was technical.

I have less experience in this area than with data analysis in Python. But I’ve also separately been building some tools to pull US government economic data through APIs (and also working on some ETL functions).

Earned: Amazon AWS cloud practitioner crrtification

I also passed the AWS cloud practitioner certification. This was a pretty high-level overview of AWS and some of the products AWS offfers.

I wanted to do this as a jump-off to the AWS Machine Learning specialty certification. I was less than impressed with some previous AWS machine learning training I’ve done (more in that below), so I’m hoping the AWS-provided training complemented with some LinkedIn Learning courses is a better path.

Quit: Udacity AWS Machine Learning Foundations course

I received a scholarship to this class, and it just wasn’t cutting it for me. There was minimal background around the kinds of AWS machine learning tools I’d use professionally. All I want is a couple labs to learn Amazon SageMaker and a sandbox environment to test some basic models.

The Python part of the course was a good refresher though.

Overall I’m pretty disappointed. And it makes me seriously question actually spending money on other Udacity courses. I was interested in the Data Engineering course, but I’ll almost certainly take the AWS or Azure or GCP equivalent instead.

Next up

AWS Machine Learning specialty certification

Like I said I jumped into AWS to prepare for the AWS Certified Machine Learning certification immediately after finishing the cloud practitioner certification. I’d like to finish this in the next six-to-eight weeks, so I’ll put a soft deadline of Thanksgiving on this one.

But I may have to push that back a bit because I’m starting…

Build a Web3 App on Ethereum with Solidity + Smart Contracts

I own literally zero crypto. And aside from a brief period of buying bitcoin (around $19K) and selling about a week later (that was a mistake), I am at zero experience with anything crypto/blockchain.

I’ve been reading a lot lately about some new DeFi products though, and I think a lot of trends in the space are very interesting. And I thought what better way to learn than to build.

So this is a two-week course offered by _buildspace, starting 10-Oct.

I’m pretty jazzed about this one. Look for the comprehensive review and my own line of NFTs soon.


Writing on the internet has been good to me. Now here come some numbers and models.

Hello. Here’s my first post on my revamped personal website.

Since about 2007, I have been writing on the internet. From personal blogs to freelance gigs to writing professionally, I’ve put a lot of digital ink to digital paper over the years.

In the past few years, I’ve also been trying to build up more technical skills – data analysis in Python and SQL, machine learning with scikit-learn, and lately learning to build data pipelines and data engineering.

So where does this leave this personal blog? Probably somewhere in between.


My writing these days is behind a paywall, so I can’t exactly link to what I’m writing about day to day. So I’ll probably step back and occasionally write about bigger picture things. Writers like Josh Barro and Scott Galloway and Tracy Alloway and Allison Schrager are sort of my template here.

I’ve had several iterations of a Substack over the past year, but found that wasn’t really the model for me. More often than not, I’m writing for my own edifice – exploring my own ideas, organizing my thoughts, and learning more about something as I write about it. That’s just easier to do (for me, at least) in a good old fashioned 2008-style blog.

Data engineering and machine learning

I’m also hoping to share some of my data projects, and updates on my path as I dive deeper into my foray into analytics, machine learning, and data engineering. I worked for several years as a data analyst, but there’s a wide, wide world out there that I want to explore.

So what am I doing right now?


AWS Cloud Practitioner Essentials

I’m doing this as a stepping stone to next do either:

I’m leaning toward the latter, though I would guess I’d have more professional use for the former. Let’s get through the Essentials training, and we’ll see where that goes.

I also have more to write about how I got to this point, and why I’m not focusing on Azure or GCP. Look out for another post soon.

Udacity: AWS Machine Learning Fundamentals

I have some thoughts on this course. First, I received a scholarship for this nanodegree program. The top 400 performers then have the chance to earn a scholarship for the AWS Machine Learning Engineer nanodegree program.

But so far, I’m incredibly disappointed by the AWS Machine Learning Fundamentals program. It’s just a quick rundown for some of the AWS machine learning tools that are available – but doesn’t include any labs or really anything professionally applicable to learning machine learning on the platform. There’s some Python coding on it, but another free Udacity program I’ve previously completed – Introduction to Python Programming – is much more valuable, even as somebody who has been coding in Python for a couple of years.

I’m not sure I’ll spend much more time on this program, and I certainly won’t bother trying to hard to get Udacity’s Machine Learning Engineer scholarship. The program provided directly by AWS (Machine Learning – Specialty) seems to be much more thorough – plus I’d much rather work to get AWS’s certification rather than Udacity’s.

Outside of those trainings, I’ve just completed my company’s Python certification, and I am working toward our internal API certification.

The path ahead

Here’s an incomplete list of what I’d like to improve/learn in the months ahead:

  • Git
  • Looker
  • Airflow
  • Dbt
  • Docker
  • PyTorch
  • AutoML

Project Portfolio

Here’s a link to my GitHub. I need to update each of the projects listed there, as they’ve all seen fairly significant updates since I last pushed my code there. I also have a few more projects that I’m hoping to put up there soon, so consider my linking to it as a way to publicly put pressure on myself to actually get to it sooner than later.