I was fighting with a home automation this morning and it required me to ping a few high-traffic websites and I got curious looking at the headers that Apple returns:

$ curl -I https://www.apple.com/
HTTP/2 200
server: Apple
content-type: text/html; charset=utf-8
x-frame-options: SAMEORIGIN
x-xss-protection: 1; mode=block
x-content-type-options: nosniff
strict-transport-security: max-age=31536000; includeSubdomains; preload
referrer-policy: no-referrer-when-downgrade
content-security-policy: default-src 'self' blob: data: *.akamaized.net *.apple.com *.apple-mapkit.com *.cdn-apple.com *.organicfruitapps.com; child-src blob: embed.music.apple.com embed.podcasts.apple.com https://recyclingprogram.apple.com swdlp.apple.com www.apple.com www.instagram.com platform.twitter.com www.youtube-nocookie.com; img-src 'unsafe-inline' blob: data: *.apple.com *.apple-mapkit.com *.cdn-apple.com *.mzstatic.com; script-src 'unsafe-inline' 'unsafe-eval' blob: *.apple.com *.apple-mapkit.com www.instagram.com platform.twitter.com; style-src 'unsafe-inline' *.apple.com
cache-control: max-age=582
expires: Sat, 23 Dec 2023 12:49:53 GMT
date: Sat, 23 Dec 2023 12:40:11 GMT
x-cache: TCP_MEM_HIT from a23-218-251-35.deploy.akamaitechnologies.com (AkamaiGHost/11.3.3-52660090) (-)
set-cookie: geo=US; path=/; domain=.apple.com

What happens next will shock you…

As somebody who's spent the majority of his life figuring out how to make computers do what I want by carefully coaxing out the one-and-only correct commands in the one-and-only correct order, the relative chaos of figuring out what works and what doesn't to get LLMs like GPT-4 to do what I want has really pushed me out of my comfort zone.

Case-in-point, I was working on modifying a GPT script to improve the grammar of Japanese text—something I can fire off with a Raycast script command to proofread my text messages before I hit send.

I'd written all the code to talk to the OpenAI API. I'd sent a prompt to the computer to fix any mistakes in the text. It should have just worked.

Instead, running the script with a prompt like this:

Content warning: more content…

One thing I've been thinking about for a while is the ability of LLMs to generate convincing-enough-for-somebody-to-run fake leaks, so I decided to try my hand at creating a fake Apple leak.

I'm too cheap to pay for ChatGPT+ and I'm too lazy to have done this via OpenAI's GPT-4 API, so I used Bing Chat to take a crack at generating the Apple Newsroom article that we'll see go up on June 5th after Apple announces its headset. I figured that there was so many recent rumor round-ups and so much prior art of Apple Newsroom articles that it'd probably do a halfway decent job of this.

Here was my prompt:

Write me a press release in the style of an Apple Newsroom story announcing the new Apple Reality Pro headset that's based on only the most recent set of Rumors from Mark Gurman and MacRumors

I thought its response wasn't half-bad:

Turns out, there's more to it…

When I first pitched doing a RubyKaigi "Field Report" summary for Test Double, my wonderful and benevolent boss Cathy Colliver suggested that I make a short video explaining how I pack so light for each trip. It took me a few weeks to get around to it, but here it is!

I've been living with this setup for about 6 years now and have barely changed a thing. If there's a headline to be had here, it's that the less you carry, the less you have to worry about.

If you'd like to try out this lifestyle, here are some links of stuff in the video:

What happens next will shock you…

Over the last three weeks, I've been playing with Bing Chat as a sort of international travel and language advisor, and I've learned a few interesting things along the way that may help others get more creative with how they use AI.

There are three main categories where I've found Bing Chat helpful:

• Generating ideas of where to go and what to do next
• Translating phrases that straddle nuanced cultural differences dictionaries and translation software can
• Answering "why" questions that would normally require a human

There's also one "gotcha" that's particularly interesting, but I'll cover that at the end.

You'll never guess what happens next…

Today I'd like to tell you about a very special place that not very many people will ever get the opportunity to visit.

After concluding my duties as a field reporter of RubyKaigi 2023, I found myself with a luxurious seven or eight days to myself. While at Kaigi, I asked several Japanese friends where I should visit in the southern island of Kyushu. Almost to a person, they said "Kagoshima".

So, I went. And it was great! The weather was warm, the nature was beautiful, and the people were easygoing.

Immediately, I wanted to go deeper. I got it in my head that if I went somewhere even more inaka I could prove I was a Real Vlogger by going on an adventure and then making a YouTube Short set to an epic piano score.

But… where to?

To be continued…

I became familiar with Blue Ocean Strategy in the context of Nintendo's decision to forego the "console wars".

Instead of pushing to design consoles with the fastest chips and best graphics, they embarked on quirky industrial designs and user experiences with seeming tangents like the Nintendo DS, the Wii, the 3DS, the Wii U, and the Switch. The big idea (as it was baby-birded to me via amateur videogame journalism) was that competing with Sony and Microsoft to have the highest-performance machine or the best-looking version of multi-platform games was a losing proposition. It represented competing in a so-called "red ocean" (as in, there's blood in the water). Why? Because there were already two well-funded competitors vying to sell the exact same thing. The best Nintendo could hope to do was be marginally better, despite being in a far weaker capital position, with less access to the top chipmakers, and with a stable of IP that didn't necessarily benefit as directly from better graphics. So they pursued a blue ocean strategy by creating bold (and occassionally bizarre) products that couldn't be compared apples-to-apples with the competition.

The rapid, unscheduled disassembly of Twitter-dot-com over the last six months has resulted in an ocean of opportunity emerging. Tons of entrants are getting in. Mastodon was already there. Hive was there for a couple weeks, too. Journalists toyed with giving up the one thing they really care about—drip-feed dopamine from constant notifications—to join Post.news. Meta is apparently building a Twitter-like platform with ActivityPub support. Spoutible is a thing, I guess. And this week, everyone's talking about Bluesky, the open and federated but nevertheless locked-down and invite-only Twitter clone that started under Jack Dorsey Twitter and whose new app looks more like Twitter than Twitter does.

All these real-time, text-based activity streams are pouring chum straight into a deep red ocean.

And before you knew it…

…You're right, it shouldn't be this hard

I'll never forget when I bought the first 5K Retina iMac. Almost as soon as I ripped it out of the box, I booted it while holding down Command-F2, assuming it would support Target Display Mode, with the intention of using its one-of-a-kind display with my gaming PC. I was heartbroken when Mac OS X booted anyway and I slowly realized that Target Display Mode hadn't survived the transition to retina resolutions. And it never came back, either. (I haven't really been happy with my setup ever since.)

Well, here we are, 8 years later and Apple has introduced the 5K Studio Display. I ordered one the minute that they hit the store in the hope I would receive what I thought I had purchased in 2014: a single 5K Apple display that could drive both a Mac and a PC desktop. (Nevermind the fact that it's damn near the exact same panel that I bought 8 years ago.) When my Studio Display arrived, I tore it out of its environmentally-friendly origami box and excitedly plugged it into one of my Nvidia RTX 3090's DisplayPorts with a DisplayPort-to-USB-C cable.

I booted up the PC: nothing happened.

Let's dive in and find out…

…without actually doing that

How we got here

After Apple released its (soon-to-be) previous generation Mac Pro, it probably didn't take long for them to realize they had a trash can fire on their hands, especially with regards to GPU performance. When Apple announced eGPU support for macOS in 2017's High Sierra release, it was hard not to see the announcement as anything more than an admission that Apple's top-of-the-line desktops and notebooks shipped with subpar GPUs due to their severe thermal constraints. Of course, because Apple has never considered AAA gaming to be an important function of its products, the Mac has always lagged behind Windows in GPU availability and support. But by 2017 (and until the new Mac Pro tower releases this fall), the situation has been especially grim: even for workstation tasks like video encoding and 3D modeling, the internal GPUs Apple has been selling are so bad that they're driving a nontrivial number of creative professionals—a market Apple actually does care about—off its platform.

The world may be excited to close the door on the ill-conceived trash can Mac Pro, but if it hadn't been for its glaring design flaws, Apple and Intel probably wouldn't have prioritized the engineering needed to make running an eGPU over Thunderbolt 3 a commercial reality.

You'll never guess what happens next…

For years, economists have been puzzling over why, despite unprecedented technological innovation since the dawn of the Internet, productivity is flat. Really, nobody seems to know why! Look no further than this week's news to find a consensus opinion that the just-around-the-corner cure for lagging productivity numbers is—wait for it—more technological innovation.

Productivity is a curiously-named economic measure that essentially boils down to "amount of money you generate for your employer over time." And because the promise of most technology is to enable people to do work faster, we should expect technology's useful impact to be measurable, even with an (oversimplified) equation like Labor + Technology = Productivity.

But something has clearly gone wrong. If we work backwards, we already know productivity is flat. And we are equally certain that technology has improved over the last twenty years. That leaves just one variable for which a negative value could explain the productivity gap: maybe we're literally doing less useful work every day. Reflecting on my own experience, I'd go a step further and ask, what if recent technological advances are actually decreasing our productivity?

But wait, there's more…

Here it is, the post that enumerates all of the ways in which remote work has turned me into a total weirdo.

For almost a decade now I've been working from home, enjoying the unusual freedoms—and anxieties—that it brings. If a single theme has emerged, it's this: by default, I'm an undisciplined mess. When given the choice between short-term distractions and long-term goals, I'll take the passing hit of instant gratification every time. (This paragraph took me ten minutes to write because I was text messaging back-and-forth with @hone02.)

Only one thing can overcome my lack of self control: replace all my good intentions with hard and fast rules, then stick to them so rigidly that my constant fear of failure will inadvertently be put to productive use.

To be continued…

[A translation of this post is available in Chinese and in Spanish ]

Programmers often describe their ideal tools with adjectives like "powerful", "feature-rich", and "highly-configurable". Few users are seen as wanting more from their computers than programmers.

This popular notion agrees with our general intuition that more capability intrinsically yields greater productivity. My lived experience suggests, however, that while capability is a prerequisite for productivity, the two hardly share a linear correlation. A dozen ways to do the same thing just results in time-wasting analysis paralysis. Apps packed with features to cover every conceivable need will slowly crowd out the tool's primary use. Every extra configuration option that I delight in tweaking is another if-else branch in the system, requiring its developers to test more and change less, slowing the pace of innovation.

And then what happened?…

In seven days, it will have been ten years since the keynote to end all keynotes.

A decade hence, I'm uncomfortable with declarations that the iPhone and iOS are "mature," however. Unlike other mature platforms that became commoditized and absorbed into daily life, smartphones have not receded into the periphery of our attention. If anything, the entire planet is more glued to their phones today than ever before. "Mature" is code for "set in stone", and it's my view that any device that's able to so thoroughly capture the attention of the human race as the iPhone-era smartphone demands continuous reinvestment. Past decisions should be constantly reconsidered.

And after a year like 2016, I like to hope that Apple is reconsidering some of the decisions they've made about how their platforms have influenced life on earth, including our politics.

To be continued…

First, some background

I embarked on a spiritual journey this week to answer these questions: do all the developers I see switching (or threatening to switch) to Windows see something that I don't? Has Microsoft actually turned the Windows user experience around? Is the combination of Microsoft's hardware & software superior to Apple's on the Mac?

At the heart of these questions: hope. Hope that people have placed in Microsoft's embrace of open source, its in-house hardware design, and its broadened cross-platform support. My suspicion at the outset of this experiment was that the recent, near-universal praise of Windows is, instead, mostly hype—fueled in large part by a general frustration that Apple has been the only serious contender for developer mindshare for over a decade. Most of the people I know switching to Windows of-late are furious about Apple's apparent product direction, and I'm biased to think their praise of Windows represents a sort of motivated reasoning. But I can't test my bias without seeing Windows appraised by someone who, like me, is a genuine fan of Apple's products… reluctantly, I figured I'd do it myself.

So, last week, I bought a Surface Pro 4 from the local Microsoft Store and put it through its paces on my own terms and carrying my own biases. I'm not a dispassionate reviewer, poring over each feature checkbox and assessing its objective quality. I'm just someone who really likes using iOS and macOS, but is interested in challenging Microsoft's products by asking them to "make me switch." These are my initial notes.

Let's dive in and find out…

I'll have a lot more to say about my experiments in trying out Windows over the coming days, but as a special Christmas bonus edition, I thought I'd share the steps that were apparently required for me to register my Surface Pro 4 with Microsoft.

As I got in bed last night, I activated tablet mode for the first time and while perusing the don't-call-it-Metro tile page, I saw an app called "Surface". I have one of those, I thought, I should tap that!

At first blush, the purpose of the app is to introduce you to the Surface's features, process device registration, solicit customer feedback, and so forth. The first thing the app asks of its users is to register the Surface device for benefits that include both requesting (and cancelling!) hardware service. Since part of my aforementioned experiment is to begrudgingly click "yes" to every asanine pop-up and prompt the operating system throws at me, I decided to go ahead and register the device.

Keep reading…

I bought a Surface Pro 4 tonight. Over the next week I'm going to share my notes on getting used to it. And over the next month I'm going to attempt to do all of my open source work from it. but tonight, I want to first comment on the state of the Mac and PC so that I can get my own initial perspective and biases on the table.

Subsequent posts in this series include: this and this.

What happens next will shock you…

My wife & I have been returning to my parents house for Thanksgiving every year since we graduated college. Customarily, people stop schlepping home to exploit the free labor of their parents' holiday cheer around the time they have kids—but for a growing number of millennials who aren't interested in bringing children into the world, this presents a dilemma: do I expect my mom to cook me turkeys until she is physically unable?

Short answer: mostly, sure. [Love you, mom!]

Longer answer: this year in particular, having logged over 25 weeks traveling, I was eager to stay home for the holidays for once. I still wanted a traditionally large and wasteful Thanksgiving feast, but I just didn't have it in me to drive three hours for one.

Spoiler alert: there's more to this…

It's time we had a talk about the news.

From the first time I heard a modem chime, I've been on a mission to discover, curate, and editorialize as much news as possible. Over the years, I've written for a variety of outlets — called "web sites" at first, then "blogs" for a while, and now subsumed by the lifeless social media platforms to which we all contribute. From the mid-nineties until I graduated college, I slowly optimized this information funnel — broadening the aperture of content I could absorb each day, while tightening my own editorial voice. Being informed calmed my anxieties about the unknown world, whereas honing a distinct persona gave me a sense of control as I navigated it.

But optimizations that lack a limiting factor run the risk of becoming too successful. My voracious appetite for novel content was so far outside the norm that a developer working on Google Reader once contacted me to ask what I was accomplishing by using the service — apparently it was unusual that I'd been reading an average of 900 articles a day at an uninterrupted pace for over 7 years.

And before you knew it…

The degree to which we're constantly connected via our phones is well-documented, thoroughly-researched, and intuitively understood by over a billion people. But we tend not to do much about it, because the benefits of being connected seem far more concrete than the relatively intangible benefits of being disconnected.

This week, I was forced to reckon with this in a stark enough way to convince me to take action.

And before you knew it…

I often tell people the only reason I blog or speak is to guard against having the same conversation over and over again. So, here I go again:

• The first thing I eat every day is dinner, meaning I don't eat breakfast, lunch, or daytime snacks
• I run 3 miles every day, without exception. In spite of my travels taking me all over the world, I'm pretty sure I haven't missed a run in two years

Doing these two things has improved my health considerably, saves a lot of time and money, and has improved my ability to focus on my work.

Keep reading…