RSS

A Musing Bean

The Mac Studio Display


I held onto my 2014 27" 5K iMac (the OG) for over 7 years hoping for an M1 replacement, but alas it was not to be. Instead, Apple is pushing us to buy a Mac Studio ($2,000) and the Mac Studio Display ($1,600). That’s a total of $3,600 for the base models compared to the $1,800 I paid for the iMac. Sure, inflation has been up, but it's not been 100% in 7 years.

Of course, I could get an M1 Mac Mini ($1,100 for a 16 GB RAM + 512GB SSD model) with the display, for a cheaper $2,700 total, but that's still a lot of money.

No matter how you slice it, $1,600 for a display is a lot of money in 2022, when you can pick up any number of highly rated 4K 27" monitors for under $400.

Here’s what's irksome - in 7 years the computer industry has failed to produce a viable 5K monitor with a glossy screen (which has sharper and more vibrant colors than the usual matte). Yes, there's the LG Ultrafine, which has the same panel as the 27" iMac, for… $1,300, and that's about it.

This is an example of a market failure. If Dell had a 5K Ultrasharp display in the market today for around $800, they's make a killing. But somehow they don’t.

So, it looks like I'll have to settle for a Mac Studio (base model with a 1TB SSD) and a "cheap" 4K monitor, or two.

Related: Tech.

Success = Talent + Technology


Question: How many piano prodigies were there in the past 10 years?
Answer: Just as many as any 10 year period.

But few people know or care who they are.

Why? Because every new piano prodigy must compete with every prodigy who has ever existed.

If you leverage a new technology, then you are only competing with the other people doing so right now.

Van Gogh became famous because he leveraged modern pigmented paints, yellow in particular, which were introduced just when he was active (his brother was an art dealer who set his up).

Bill Gates famously was one of the first teenagers to get access to a computer in his (private) high school.

There is a cadre of AI / ML prodigies being minted right now.

Success is when talent meets new technology. What emerging technologies will you leverage?

Related: Business, tech.

Why other people's opinions don't matter at all


It's a cosmic accident that you happen to be alive at this particular time. You could have just have easily been born a hundred years earlier, or a hundred years later. Think about the kinds of things people believed a hundred years ago, or might believe a hundred years from now. Would you agree with much of them?

Do the things people believed a hundred years ago bother you today? Should anything anyone else believes now or in the future bother you today?

Related: Thinking.

The Stairstep Rule


When refactoring code, there's a rule to either change the structure or change the functionality, but never both at the same time.

This applies to other areas of life.

When modifying systems (especially complex ones), holding the outputs (the expected functionality) constant acts as a set of test cases.

Likewise, the structure constrains what functionality is achievable.

When you want to add functionality not possible with the current structure, you need to change the structure first.

The tendency is to want to do both at the same time, but changing the structure of a system will invariable break some existing functionality. By first changing the structure, you can more clearly assess what functionality will break, and decide if that is an acceptable tradeoff.

Related: Design.

Moving to a Surface Laptop from MacBook Air


After waiting years for Apple to release a decent laptop, I finally gave up and ordered a Surface Laptop today. It's a new 1st gen, 2017 model (i7-7660U, 8GB RAM, 256GB SSD, burgundy red) for under $800.

I'm a bit surprised at my own decision, which I think sheds a bit of light on just how badly Apple has lost their way.

On the surface (ha), my reasoning in straight forward: I want a portable laptop to work for a few hours at a time in a coffeeshop, maybe to take on trips. It should have a sharp screen, a good keyboard, 5-6 hours of real battery life, and ok performance. I'm not looking to edit videos or play heavy games on it.

The last laptop that fit the bill was the previous-gen MacBook Air, and it fit it to a T. I still use my 2012 model, which I originally purchased for $1,700. It has worked ok for the past 7 years, but using a non-retina screen in 2019 is inhumane.

I was hoping Apple would simply address that with the 2018 model - just slap on a retina screen, put in new processors, and keep everything else the same, especially the keyboard.

Setting aside the price - while obscene, I was ready to pay the Apple premium if the right product existed - the main thing that drove my decision was the keyboard.

The Butterfly Effect


It's hard to overstate how important a good keyboard is. As a software developer, it's an essential tool for me. I'd argue that the keyboard is perhaps the most important component of a PC, followed by a large and sharp screen.

A suboptimal keyboard has a direct and huge impact on my productivity.

The Apple butterfly keyboard has a double-whammy: Not only is it uncomfortable to use with comically short key travel, it continues to be plagued by reliability issues.

The real head scratcher for me is why Apple decided to make the tradeoff: between producing a marginally thinner and lighter laptop and one with a decent keyboard. There's no question that I would have ordered the first unit if the MacBook Air was a quarter pound heavier but came with the old keyboard.

Stuck in a Design Trap


I have been hoping the next design refresh would resolve these issues, but on further reflection I realize Apple has painted themselves into a corner they can't easily escape from.

Apple's engineers aren't dummies. The opposite in fact. If they could produce a good keyboard with the current constraints, they would have by now. It's been over 5 years.

If the next model is thicker and heavier, Apple will be hammered by critics, who will (rightly) see this as an admission of guilt and error, and potentially opening them up to big class-action lawsuits.

Apple's only course is to keep on denying the facts and churn out marginally thinner and lighter devices. I believe they will eventually solve the reliability issues, but I don't see how they can possibly improve the key travel without a quantum leap in space-saving technology.

Doubling down on the Wrong Path


“No matter how far you go down the wrong road, always turn back.” -- Someone wise.

How did Apple get themselves into this self-inflicted mess?

It started in 2012, when Apple released the retina MacBook, an impossibly thin and light laptop with a terrible (1st gen) keyboard. They dramatically improved it the next year, and if they had kept it at that, all would have been well.

That itself wasn't the problem, the fatal mistake was deciding to carry the butterfly keyboard over to all other models.

Even with everything I said, the retina MacBook is still a great design today - it is the thinnest and lightest notebook you can get with decent power, with the tradeoff being a poor keyboard. That's not a bad tradeoff for a device you expect to use relatively infrequently.

But it is a terrible tradeoff for a daily workhorse.

They took a design that was optimized for one particular set of constraints (thinnest and lightest possible), and moved it without enough thought to products with very different constraints.

Note that when the previous-gen “chiclet-style" keyboard was introduced, it went the other way: First appearing in the MacBook Pro, then carried to the MacBook Air. Recall also that the first-gen MacBook Air compromised on battery life and power, but not on ergonomics.

Creativity Loves (the right) Constraints


The work of design is to make the best tradeoffs given a set of constraints. Some constraints are the goals you want to achieve (“make the thinnest laptop possible” or “make the most powerful laptop possible"), and others are simply dictated by available technology (“lithium-ion batteries take up X cubic mm per watt”).

Here's the lesson: It's critical to define the right constraints before undertaking a design.

And just as valuable: It's important to revisit and question the (often unspoken and implicit) constraints imposed on a design.

The worst thing you can do is to solve the wrong problem perfectly.

Related: Design, Business, Tech.

Amazon's Echo Dot Kids Edition


This is both the most remarkable and most unremarkable thing I've seen recently. Unremarkable because is an obvious line extension. I thought Amazon already sold a kids edition of the Echo.

Remarkable because this is obviously the future, here today.

I predict that the impact of this on humanity will be far greater than the impact of all previous technology.

I say this without hyperbole.

Imagine kids growing up everywhere with a magical genie that responds to voice commands, their voice commands. Answering all their questions, and bringing them anything they ask for -- anything in audio (and soon visual, someday physical) form.

A child could use this before they even learn to read, and have access to the sum total of human knowledge, and any capabilities granted to them (by their parent's credit cards).

Imagine being someone who grows up in a world where it is the most natural thing to speak your heart's desire at any time, and have it automatically fulfilled by powers unseen.

How will people growing up with this tool be different from those who grew up without it?

How will civilization be different when everyone grows up with this tool?

Can we even think of humanity the same way once everyone is thoroughly dependent on such tools?

Are we inevitably destined to become a bio-tech symbiont?

Can there be any alternative?

Related: Society, Tech.

Renaming file extensions in Mac OSX Terminal


I had a lot of .jpeg files that I wanted to rename to .jpg. No big deal I thought, just run “mv *.jpeg *.jpg”. Nope, that doesn't work.

According to Stack Overflow, the way to do this is: “for old in *.jpeg; do mv $old `basename $old .jpeg`.jpg; done”.

Forget that.

I ended up installing and using the ren script suggested in the same thread. And even that isn't as clean as the old DOS command.

I'm amazed that this is still a thing in 2018.

Related: Code.

Hosting Static Websites on Amazon AWS S3


In the past, I've hosted this blog on site44.com, which works great, is easy to use, and fairly reasonable at $4.95 / mo. However, I wanted to experiment with even cheaper hosting options, which led me to try Amazon S3's static website hosting.

Following the steps on Kyle Galbraith’s excellent article, it took me less than 15 minutes to set up. And it works like a charm!

You can't afford NOT to own your content


I expect S3 hosting to cost less than $20 a year, possibly much less. This is 70% cheaper than running a private VM.

At this price point, there isn't a good reason to publish your content on anyone else's domain. Think about it this way: If you spend more than 2-3 hours a year writing, your time itself is worth far more than the cost of hosting, including domain registration.

At a minimum, $20 a year is worth keeping control over your content and privacy. And if your content does eventually take off, you can scale up and do what you want with it.

It's easier than it looks


The AWS console is much, much more cryptic and technical than it needs to be — reflecting a dev-focused culture. This is a shame, because the tools and processes themselves are very straight-forward. AWS presents a lot of options and flexibility, most of it you can safely ignore for hosting a simple site.

The good news is that there are lots of articles, like the one I used, that walks you through the process, and you only need to do it once.

Related: Tech.

Knowledge has a Half-Life


Think of all the knowledge and skill it took to navigate a sailing ship in the 1600s to the other side of the world. It is the equivalent of piloting a spaceship to the moon today. Worth many fortunes in its day. Yet, that enormous trove of knowledge is completely obsolete now. It has no economic or social value.

Before around the 1970s, people in technical fields who mastered advanced algebra held an advantage over their peers. But with the widespread use of electronic calculators, those set of skills became nothing more than a curiosity.

Up until the early 2000s, we regarded people who could remember vast amounts of "general knowledge” as highly intelligent. Today, you can look up details on even the most insignificant issue within moments using tools like Google. People who spend their time remembering vast amounts of trivia are at best eccentrics (or contestants on Jeopardy).

Today, society puts a high value on people who have the skills to program computers. Knowledge of current computer languages like Javascript, Swift, and Python command a premium in the job market. No one expects the same premium 20 years from now.

What's the pattern?

Technical skills and knowledge are highly valuable for a while - when new technology requires that skill. After a while, the supply of people with those skills exceeds the demand, and eventually new technology makes the previous generation of skills obsolete.

The next major technological shift is shaping up to be domain-general AI - computers that can substitute for a human (or enhance humans) in narrow domains. There will be a tremendous demand for the technical skills to design, produce, and manage these AIs.

Three Learning LifeHacks from Marc Andreessen


Lots of great content in this a16z Podcast where Don Faul interviews Marc Andreessen, but 3 unexpected learning hacks pop up right at the end:
  1. Stick a bluetooth headset in your ear to listed to podcasts and audiobooks all day.
  2. Use the iPhone's text-to-speech tool to read any web page to you, such as Wikipedia pages.
  3. Research papers are under-exploited. Find the top professor in a field, and read all their published papers.

Related: Business.

Older Posts