Reminds me of a bank account I had in Australia back in 1999 (I think it was Commonwealth Bank). I'd opened it as I was there travelling for a year and it made EFTPOS (early payment by card) easier and cheaper than using my UK bank card.
At the end of the year of travel, I went into the branch to close the account and they said I couldn't close it because I couldn't withdraw the remaining $2.11 because it was lower than the minimum withdrawal amount (I think maybe $10) and that I should just leave it open.
About a year later, I got the latest 6 month statement that they'd posted internationally saying that my account was now $100 ish overdrawn, because they'd started charging me a monthly fee on the account, and somehow I was just expected to have known that. The previous statement hadn't mentioned it, they'd just started charging it because they'd introduced a fee for accounts with a low balance. It took about several long international phone calls (and back then they were about $0.40 per minute, at least one was probably an hour long) to convince them that I had no plans to return to Australia in the foreseeable future, or to pay them the account fee especially given that I had attempted to close the account previously and they refused, and finally they agreed to close the account and waive the fee. But I'd spent loads in charges on the international calls, and they'd posted me a statement twice printed on heavy non-airmail paper, all because they wouldn't just let me shut my account when I originally asked in branch to do it.
25 years later and I still haven't gone back to Australia, even though it was one of the best years of my life. I'm going blame the bank for that (even though it was really just not having the opportunity to go again!)
I assume his point is that making stuff that assumes a mouse makes for a bad text-based UI. Absolutely fine if everything is controllable via the keyboard, e.g. if the tabs were labelled F1-Fn and they function keys switched them, or they had an underlined letter and Ctrl+letter switched focus to it, or whatever.
But if this thing requires you to just tab a lot through lots of pointless and rarely used fields to get to a "button" so you can activate it, because it's really all designed to be used with a mouse, then it's a bad text-based UI.
There are some incredibly good text-based UIs around, some going back to mainframe stuff from the 70s. Most of them are optimised for speed of control via keyboard rather than for looking pretty. Almost none of them would be quicker to use with a mouse.
From the follow-up article: https://rachelbythebay.com/w/2022/01/27/scale/
"If you are sticking with the 2022 equivalent of MS-DOS and machines that do exactly one thing at a time, then yeah, you are going to have a whole fleet of systems all sitting there busy-waiting on something stupid."
"Don't recreate the basement full of PCs when the problem can actually be solved with a single box sitting in a cabinet somewhere."
If they really were just PCs to act as a modem-to-network bridge, this seems to be remarkably cost-inefficient. I remember around 1997 helping the university chuck out a few serial line concentrators (no idea what they were actually called), each of which had 32x RS-232 ports that worked up to 19200 baud and a 10Mbit coax network connection at the back. The on-board computer (wouldn't be at all surprised if it was much more than a 68000 or two) interfaced with all the serial ports and translated it to telnet on a remote machine. You could also send it an escape code and then via a primitive command line connect to any arbitrary IP address and port over TCP. I remember in my student days (so maybe 1995) using finger and SMTP directly from these text terminals without actually logging on.
No idea when these became available, but we were chucking them out in 1997 or 1998 as we were upgrading the labs of text terminals to PCs, so they probably at least a decade old by then.