The primary completion, "Lisa R", is someone I have not exchanged email with in probably 15 years. "Lisa H" is someone I email multiple times a day. They're not even alphabetical!!
17 Responses:
The primary completion, "Lisa R", is someone I have not exchanged email with in probably 15 years. "Lisa H" is someone I email multiple times a day. They're not even alphabetical!!
“The primary completion” is what's probably “wrong” with your thinking. Someone at UX design may have guessed that the item closest to the on-screen keyboard is the primary one since it's easier to select. Many meeting hours have died to bring you this information.
:mind-blown-emoji:
It sounds logical now that you explain it but the cognitive dissonance going on in my 40-years-of-physical-keyboard-and-mouse-focused-UIs brain is making me dizzy.
I'd be really interested to know if someone who has primarily used phone touchscreen interfaces their whole life would find it natural to have the obvious choice at the bottom of the list.
Incorrect. If I typed space, the top one (Lisa R) is the address that was inserted.
I feel so important now.
See, now you're just encouraging it!!!
"...and that's how giving you the alias 'Hastur' on my phone resulted in the apocalypse being triggered.
Sorry."
Mine has a very bad habit of changing math into meth.
The first time was funny but it keeps happening and now I worry about why it thinks that.
My work mail client keeps autocorrecting properly spelled words and names to misspellings. The hypothesis is that we share a pooled dictionary that anyone can add new “words” into. And since we’re a bunch of engineers who are notoriously bad spellers, entropy is creeping in.
Autocorrect is even making professional journalists look bad, substituting in the wrong homonym. For example I keep reading articles mentioning that the driver hit the “breaks” too late.
Some web mail providers maintain per-domain spelling models. For example, inside a corporation that sells Viagra, they'll eventually learn Viagra is not spam, and it will be included in the dictionary. This has caused some serious problems recently within large mail providers.
That's very generous of you. Most professional journalists have editors, and any editor worth their salt would spot that a mile away.
The written form is being rapidly unlearned in more and more apparent ways. Asking questions these days has devolved into "How to wifi?" or "How to use?". And non-questions ("I'm still waiting for my food?") and almost-rights ("How to master Fortran with minimal effort?") are everywhere. Plus eye-twitchers like "but also this one, which I don't even know the meaning of it."
Does it fall within the natual evolution of language? Maybe? Barely? Probably.
How to stem the tide?
It appears as if traditional editors and proofreaders are being value engineered away in today’s pangolin eat pangolin era of new media. In their place are autocorrect and the peanut gallery of the comments section to pedantically point out errors. Speaking of which kudos to JWZ for doing a great job of self editing right here.
As for the “How to …?” becoming. The norm: we write a lot of technical articles that are ultimately published to our customers. Right before publication, they pass through an alleged “grammar review” step done by someone with no tech background but supposedly good English skills. As far as I can tell all they do is to load the article into MS Word and then right click on every blue or red squiggle to select what Micro$1 thinks is the correct grammar or spelling. That leads to articles written so stiffly and awkwardly that they’re almost unreadable. One of those changes is converting titles like “How to Initialize a Left Handed Smoke Shifter” into unwelcome questions. Another common change is conversion of esoteric tech terms into completely wrong but more common words. So I often have to re-edit the article to remove all of the commas, wordy phrases, and question marks. So now I call that stage of production the “Error insertion review”.
Aaack. I feel like how the English view American English now.
Mine keeps autocompleting and autocorrecting(!) „Jobs“ for my name. Did so just again as I entered above. Very annoying. Not learning. Zero learning. Stupid. Even my old Nokia T9 knew at some point that I didnt mean Cola when I spelled the name Anja.
It's not alphabetical, it's in order of prefix tree child nodes, which are typically stored as a linked list in insertion order and not sorted. Some systems will then sort the matching items by frecency (frequency/recency) but this one did not.
The Google Phone app on Android does not either. I have 8 "Nicolas" in my contacts list, only 1 of them is the person I call more than once per year, and his contact entry is always displayed last, forcing me to scroll down or to type his full name.
It's nudging you toward being more social, reconnecting to all those Nicolases that have been yearning for all these years to hear again from you, lonely and sad, forgotten. All of Google's AI, spinning their CPU furiously to find the correct trigger to push you into action, pinning to make you a better human. The problem is not the UI, the algorithm, the heuristic. The problem is you.
You just need to start calling the Lisa you care about 00Lisa and everything will be "fine".
I think you could have phrased that better.