What is interesting from all this is how quickly it appears that Starfield has tailed down from its impressive launch, dropping down just two months after it released; by way of comparison, Skyrim's original release had fewer players at launch in 2011, peaking at around 290,000, but didn't fall below 20,000 players until almost seven years later in May 2018. (Two months after release, Skyrim was still on somewhere around 90,000 concurrent players on Steam.)

Keep in mind, Skyrim's players are far likelier to be playing on Steam (both because it launched there and due to better mod support) than the official Xbox app, where Game Pass subscribers have always had "free" access to Starfield.

That said, it's hard to read this without thinking about the chorus of anecdotal reports I've heard from people who've played Starfield: Bethesda successfully delivered another "one of those" games, but it simply lacks the magic and awe that gave Skyrim its staying power.

Personally, I think a lot of this is due to the game's setting. High fantasy lends itself to dramatic moments celebrated by epic orchestral movements in a way that feels tacked on when applied to a sterile sci-fi environment. Fast travel between star systems in Starfield undercuts its sense of scale to a degree that manages to make it feel smaller than Skyrim's technically-not-quite-as-big-in-aggregate map.

These combine to rob Starfield of what industry folks call "emergent gameplay." A random hail from a starship feels like a procedurally-generated interruption, whereas a fellow traveler approaching you along a trail at night feels like a natural coincidence or kismet. Both might be random scripted events, but only one maintains suspension of disbelief. Similarly, knowing that I could walk thirty minutes and eventually reach a mountain in the center of Skyrim's map lends credibility to its environment, but being forced to constantly fast travel between planets shatters any illusion that Starfield players are exploring a single connected universe.

As a result, Starfield is merely a good game. It's more Fallout than Elder Scrolls. And where Fallout oozes with charm and satire, Starfield feels flat and mundane. The premise may be more realistic, but the world feels less real as a result.

Finishing Starfield's campaign and then deciding to continue exploring its world raises the question, "why?" It would be like checking out at Costco and then deciding to walk another lap through the store. Maybe there was something you didn't check off your list the first time around, but one imagines you're going to impatiently hustle to get it and get out.

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:

Let's dive in and find out…

While nothing has yet been confirmed, it appears people were able to figure out (or someone accidentally shared) Emergency Pizza codes that could be used over and over again by the same customer. This, obviously, isn't how the program was intended to work.

Expensive bug.

During the M1 generation of Apple Silicon Macs, one area of controversy was whether the default 8GB of RAM was less of a performance bottleneck as it was under Intel Macs. The pundits contended that a paltry 8GB was unacceptably stingy on Apple's part, but in practice—and who's to say whether it had to do with the ARM CPU or the unified memory architecture or the blazing-fast SSD performance—no matter how much you threw at the M1 Macs, it never seemed to get bogged down by swapping out to disk.

Well, that era is apparently over, because the new M3 with 8GB is getting absolutely smoked by the same chip when it has 16GB to work with:

The 8GB model suffered double-digit losses in Cinebench benchmarks, and took several minutes longer to complete photo-merging jobs in Photoshop as well as media exports in Final Cut and Adobe Lightroom Classic.

If you don't watch the video, it's worth clicking through to MacRumors' summary for the charts alone.

Big day at work. Emphases mine:

Test Double's mission is to improve the way the world builds software. We've done that by building great teams and great software—with a special focus on building things right.

Pathfinder Product's mission is to unleash greatness through modern product management. They've done that by being passionate problem solvers—with a special focus on building the right thing.

That about sums it up. When you put the two teams side by side, it's uncanny how good a fit they are. Each has hired top practitioners in their field, united by an insatiable drive to make this broken, messy world of software work better for everybody.

And as much as I bristle at the word "synergy", it's really there in this case: brilliant product strategy goes nowhere without execution, and high-performance delivery is a waste of money if it drives you in the wrong direction. The real beneficiaries of that synergy, though, will be the clients that entrust Test Double to help them accomplish both.

Apple PR representative Starlayne Meza confirmed the company's plans to The Verge. The company encourages those who have been holding out hope for a larger iMac to consider the Studio Display and Mac Studio or Mac mini, which pair a 27-inch 5K screen with a separate computer, compared to the all-in-one design of the iMac.

In the post-Jobs era, has Apple PR ever confirmed they had no plans to create a particular product like this?

Last year, I was featured on the season 1 finale of Matt Swanson's excellent YAGNI podcast. Well, season 2 kicked off yesterday with an interview with Charity Majors and as one of the few programming podcast series I can stand to listen to (high praise, trust me), I recommend you or the programmer in your life consider subscribing!

I'd been thinking about my appearance last year on YAGNI because in his lead-in, Matt joked that he had cajoled me into writing my own test framework as an alternative to RSpec. I have no recollection of this, but he turned out to have succeeded, because 9 months later I indeed published the definitive Ruby testing framework, TLDR.

Remote workers will sign off at 1 pm to participate in a digital rally and picket via Zoom.

That's sure to get their attention.