Patrick McKenzie, aka patio11, used a LLM to generate and run a DnD campaign for a gaming convention/conference, He discusses the process and results as part of this podcast on the con.
Unfortunately, in many organizations, "the library we use doesn't follow this recommendation" is a valid compelling reason. Which means that in practice "SHOULD" effectively means "WOULD BE NICE IF".
Second reason it didn't take off faster was that iPhone was typically a carrier exclusive, and in most markets, the iPhone carrier was typically one of the smaller carriers. So iPhone wasn't available to most mobile phone users in a given market, unless they went to the trouble of switching carriers.
FWIW the linked article is reporting on a 5 minute YouTube video montage of Jonathan Blow interview snippets. Might be worth it to watch/listen to the video directly.
I remember seeing a Vector Graphics computer at a computer store around 1978, when I was shopping for my first computer. I was excited by the name Vector Graphics, only to be disappointed to learn that it was a meaningless name, and their computers had nothing to do with vectors or graphics. I vaguely remember that it was a generic business machine (maybe with a 16 bit version?) with nothing to recommend it to a hobbyist over the competition.
In that era Apple had an enormous lead in graphics, software, and peripheral cards.
The web could have plausibly existed as early as FTP did. Which would have been 1972. Plenty of documents from that era had URL-like manual links of the form "pub/foo/bar.txt at MIT-AI", and many FTP servers supported anonymous login and
were fast enough for real-time text document retrieval.
It is kind of embarrassing that it took 20 years to invent URLs and browsers.
Just as a point of interest, AIUI the names "Schumpeter" and "Babbage" are "house names", not the real names of the people who write the columns. Any given column might be written by anyone, and it would still use the same house name.
Shumpeter and Babbage are the last names of famous people in fields related to the columns topics.
The article starts, "Nestled in the hills of Mountain View", which is worrisome, because as anyone with even a passing familiarity with Silicon Valley geography knows, Mountain View is mostly flat. And Google's Mountain View offices are located on reclaimed baylands, which are especially flat.
I wonder if the article authors were thinking of Xerox PARC's offices in Palo Alto's hills? Or the road named "Sand Hill Road" that used to have some venture capitalist offices? Other than those offices I can't think of any significant hill-based offices in Silicon Valley.
Supposedly Stable Diffusion was trained on 512 x 512 images, so it's not clear that it will work well for larger images even if you had the RAM.
To generate larger images, it is standard practice to generate 512 x 512 and then use a separate tool to upscale, and maybe a second separate tool to improve the face. The Windows versions of SD environments are starting to incorporate these additional tools, but the Apple Silicon versions of SD environments are lagging behind due to Pytorch metal limitations.... It'll hopefully sort itself out in the next few months.
It's likely that a significant fraction of the perf difference between Apple' GPUs and NVIDIA GPUs is due to NVIDIA's CUDA being high optimized, and Pytorch being tuned to work with CUDA.
If Pytorch's metal support improves and Apple's Metal drivers improve (big ifs), it's likely that Apple's GPUs will perform better relatively to NVIDIA than they currently do.
That's probably due to swapping due to the 8GB of RAM. People who have run Stable Diffusion on M2 airs with 16 GB of RAM seem to get performance that is in line with their GPU core count.
Some engineer (I forget if it was an Apple engineer or a Handspring engineer) once remarked that Graffiti was a much smarter way to do handwriting recognition than Newton, because: "If the Palm Pilot made a mistake, people blamed themselves for not writing Graffiti correctly. If Newton made a mistake, people blamed Apple."
I remember that shadow! FWIW it was just one line in smaller Trinitrons, and it was located 1/3rd of the way from the top or bottom edge rather than in the middle.
Apple mounted the Trinitron tube upside down compared to other vendors, so that the faint horizontal line would be in the bottom third of the screen rather than the top third.
As an outsider, I believe Rareware's secret sauce was a combination of a can-do, down-to-the-metal, fast-feedback-loop game development style that came from the pre-PC British bedroom game coders, a management team that understood how to manage game development and releases, and Nintendo's coaching on mascot development and general game polishing.
Rareware's talents were big advantages in the early 3D game console era. But by the PS2 / Xbox era, their special skills didn't help as much.
Today I'd say that Epic's Fortnite is the spiritual successor of the old Rareware.
Waymo is trialing in several northeast cities. Search for "waymo trial boston" or "waymo trial nyc".
But beyond the technical issues, there are also political issues. Search for "new york bans waymo".