If a user holds an ice cream cone upside-down and their ice cream falls to the floor, do you blame the user for not holding their ice cream cone upright or the creator of the ice cream cone for a stupid design that allows the ice cream to so easily fall out of the ice cream holding device and onto the floor?
I find far more often that bad UX is the result of someone trying to use a tool for something it wasn't designed for. They might even clob several different tools together in an unholy abomination to get it to do what they actually want instead of having a tool built to do precisely what they want (and once that tool has been built - people will inevitably misuse it to do things other than what it was designed for and then complain about its poor UX for doing those things).
> If a user holds an ice cream cone upside-down and their ice cream falls to the floor, do you blame the user for not holding their ice cream cone upright or the creator of the ice cream cone for a stupid design that allows the ice cream to so easily fall out of the ice cream holding device and onto the floor?
Playing along with this analogy, what I think we see a lot of in product development is the customer going to the PM and saying they need the cone to have a cover. The PM and the customer iterate over the specifics of the cover. PM goes to the engineer and tells them the cone needs a cover that meets x, y, and z requirements. Engineer, knowing how you're supposed to use the ice cream cone, objects. PM, knowing what the customer needs, insists.
> Engineer, knowing how you're supposed to use the ice cream cone, objects. PM, knowing what the customer needs, insists.
That’s the kicker. They know what the customer _wants_ not what they _need_
The number of times a Jr engineer has asked me “how do I accomplish task X in technology Y, it’s really important!”
I always, always ask “what problem are you trying to solve. Not once in over 15 YoE has the solution been to use X.
A good PM doesn’t say “this is what the customer needs” because most of the time the fucking customer doesn’t actually understand what they need.
The engineer knows that holding the ice cream cone upside down means they’re trying to use the product in a way it was never intended, so they push back.
A good PM would ask “why do you want to hold the ice cream cone upside down, customer?”
“Oh well we don’t actually want to hold it upside down, we just get frustrated that sometimes when we put too much ice cream on/in the cone it falls out. So if you can make the cone hold the ice cream while upside down, the problem is solved!”
“Oh, so what you actually need is a bigger cone that can hold more ice cream?”
“Oh, yeah that would work too”
meme about Spider-Man facepalming, where Spider-Man is the engineer
My point is that by letting the customer define the solution rather than explain the problem, nobody knows they're even trying to hold it upside down. The why of it all is lost in some PM / Engineer power struggle that usually results in ice cream cones having covers (and a happy customer).
> I find far more often that bad UX is the result of someone trying to use a tool for something it wasn't designed for.
Isn't that the point? In the story the engineers weren't designing a tool well-suited for the customers, but for whatever abstract scenarios they had in their head. In the open source world it's more reasonable and common to design a tool not predicated on the predominant models and workflows. And every once in a awhile those experiments result in something very valuable that helps to break predominant paradigms. But in the commercial space solving customer's immediate problems in a manner that is intuitive for them is paramount.
> In the story the engineers weren't designing a tool well-suited for the customers, but for whatever abstract scenarios they had in their head.
A joke exists about how developers will never be displaced by AI because that would require clients and/or project managers to accurately describe what they want the AI to build. On one hand that is extremely egotistical of developers. On the other it is also factual.
To my understanding of the story the developers had designed what was being communicated to them by someone who described what customers asked for and not what the customers actually wanted or needed. Nothing to do with what the engineers thought customers wanted and everything to do with what project managers had expressed to the engineers about what customers wanted. Speaking with the customers directly gave them a better idea of what was actually being asked for. So they built that instead.
My takeaway from the story is to fire the project manager. Not to make devs call clients.
Product and project managers can have similar issues of perspective as engineers, chasing trends in the industry or losing the forest for the trees (feature checklists, etc). But, yeah, having devs talk directly to customers can be problematic. If you give a customer a direct line to an engineer, sometimes they can monopolize the engineer's time (they feel they can leverage an "inside" contact) or skew their priorities. Having engineers work closely with tech support is a good middle ground, such as by fostering 1:1 relationships with support or even fielding tech support tickets themselves (tech support tickets have more finality than, e.g. sales support issues). That way they can get a broader perspective on pain points and potential new directions. It can really help refine a product beyond what can be achieved by lobbing feature and bug tickets over a fence or through formal group meetings (which often can be either too structured or too chaotic).
The argument of likeness seems odd to me because there are at least a dozen people who might look almost exactly like you who are alive somewhere in the world.
What if I give explicit permission to use my likeness but my lookalike demands it can't be used? We're both dead. Do my wishes not get respected because someone who looks like they could be my identical twin had other wishes? Whoever's estate has the deeper pockets?
See photography by François Brunelle. The similarities went past appearances too. Many of the stranger dopplegangers had similar hobbies and even similar personalities. So if an AI recreation looks like me, acts like me, and has the same hobbies as me that means nothing unless someone is trying to claim it is me (rather my likeness).
To lawyers, “likeness”is a term of art and means much more than just how you look, including image, name, voice, and other identifiable features. Basically it’s what actors bring to contracts in addition to their labor.
I don’t claim to understand all the intricacies but it is the relevant term of art when discussing this topic from a legal perspective.
Yes, but none of that is truly unique. The odds of a doppleganger sharing my name are astronomically slim but my looks, voice, interests, personality, etc. are not truly unique to me.
For an example, what of voice impersonators? Sounding like Morgan Freeman is not unique to Morgan Freeman. What if a soundalike legally changes their name to Morgan Freeman? What if a lookalike changes their name?
I'm familiar with the existence of such laws but less so with how they are enforced or how they can even be enforced at all. The laws have never made that much sense to me.
The way it mostly works is that if a company hires a Morgan Freeman impersonator to do the voice over for their car commercial, Morgan Freeman can sue them for using his likeness without permission:
> We need not and do not go so far as to hold that every imitation of a voice to advertise merchandise is actionable. We hold only that when a distinctive voice of a professional singer is widely known and is deliberately imitated in order to sell a product, the sellers have appropriated what is not theirs and have committed a tort in California.
If you not impersonating Morgan Freeman, and there is no history or framing to make that connection, he is unlikely to sue you, or win a lawsuit.
Courts are quite aware that similar things can come from divergent sources.
Even copyright law fails to protect commonality between works that would be illegal if actually copied, but were arrived at legitimately and credibly independently.
Not saying someone shouldn't use common sense to avoid problems. Don't do Morgan Freeman impersonations comedically, then voice overs of movies.
The goal I think is to protect people with valuable likenesses being undercut by cheap knockoffs, not to protect everyone's likeness. So it's more like a trademark than anything.
Very useful for flagging any O(n^2) that make assumptions about the size of N because N is not expected to exceed a certain size. Especially for when N inevitably exceeds that size.
Documenting it saves the poor dev doing profiling in the future a bit of effort so they can come up with the better solution that you failed to come up with when writing the code.
Often times code has to be written and committed and I don't have the time nor the brains to come up with a novel solution that solves a future performance issue that is not yet and is not expected to ever become a performance issue.
Rust is like walking across a mine field with all the mines flagged for you. You can dig up the mine and remove the flags over time. Or at least know to avoid stepping carelessly around them.
In C you only see the flags people know of or remembered to plant. There's an awful lot of mines left unflagged and sometimes you step on them.
It's very obvious to me which would be more safe and I find myself questioning why it is isn't so obvious to others.
I'm a Memrise beta member w/ lifetime premium access for my contributions to the site in its early days. I cannot recommend anyone use Memrise for anything nowadays it has been so heavily enshittified. In fact, I recommend against using it in favor of Anki (Memrise's biggest strength over Anki in the early days was the community mnemonics and courses (Anki equivalent "community decks") - none of which really exist in any way today).
I tried following the modern Japanese track on Memrise and was appalled at how bad it is nowadays.
After wasting the better part of a decade on speed reading as a teen and using speed reading tools I can only find myself to agree with them. Remove multiple-choice questions and ask questions about the material and speed readers comprehension crumbles apart to such a degree it is difficult to call what they do to be "reading".
There are quite a number of studies on this, but I'll reference a blog that does all the referencing for me [0] since their experience and thirst for knowledge that led them to later be an advocate against - rather than an advocate for - speed reading is basically a 1:1 match of my own.
500-600 WPM is the upper limits, 99.99% of people claiming otherwise are bullshitting, I always leave that 0.01% because some people are literally just built different and are truly one-of-a-kind (or one-of-maybe-a-dozen people on Earth). Anyone claiming such speeds is going to be under a lot of scrutiny the same way I'd be skeptical of anyone else claiming to be in the top 0.01% of anything. If someone tells me they're a Top 10 Challenger ranked League of Legends player I'm not just going to take their word for it without some solid evidence.
All that I have to say about it, is that in a place like Hacker News, you do encounter the top 0.01% on a fairly regular basis. Particularly among those who were here early on. Being too skeptical of it when you see it doesn't sound like that good an idea.
I mean seriously. Why would I lie? And why would I risk the fake reputation on this account on a lie about something stupid, when I have friends who know me here?
People can give children terrible information too and steer/groom them in harmful directions. So why stop there at "AI" or poorly defined "algorithms"?
The only content children should see is state-approved content to ensure they are only ever steered in the correct, beneficial manner to society instead of a harmful one. Anyone found trying to show minors unapproved content should be imprisoned as they are harmful to a safe society.
The type of people who groom children into violence fall under a special heading named "criminals".
Because automated systems that do the same thing lack sentience, they don't fit under this header, but this is not a good reason to allow them to reproduce harmful behaviour.
So selling the Anarchists Cookbook is illegal? Being a YouTuber targeting teens for <extreme political positions> is illegal? This is honestly news to me given how many political YouTubers there are who are apparently criminals?
Given some of the examples I'm not so sure a human would be charged for saying the same exact things the AI has said. Without an actual push to suggest violence and even that's difficult to prove in cases where it does happen (eg. The cases where people pressured others into suicide or convinced them to murder)
I would greatly appreciate if you engaged with what I wrote and not what you think I wrote if you're going to make the bold claim that I'm not engaging in good faith.
Absolutely nowhere did I equate writing a book to grooming. I equated selling the book in the greater context that "providing children with potentially harmful/dangerous information should be illegal because it grooms them to commit harmful actions to themselves or others" and this context would carry the implications that by "selling" I am referring particularly to "selling it to children" since "providing children with potentially harmful/dangerous information should be illegal because it grooms them to commit harmful actions to themselves or others". With my argument being would it be criminal for an AI but not for a human?
So to clarify the argument: Writing the book is fine. Selling the book to adults is fine. Adults reading the book is fine. But if providing dangerous information to children should be made illegal - how would selling such a book to a child not be considered illegal? Because it was written by a human and not an AI?
Every now and again a site exists that has a massive community, tons of resources, ways to speak with other learners, ways to meet language exchange partners, and are greatly successful. Then all of that gets gutted for what is essentially a worse version of Anki but for the web when the company runs out of funding and has to start turning a profit somehow. This burns the community and the people providing most of the value move elsewhere.
It's happened to italki (now iKnow), Memrise, DuoLingo, and a few sites that were so short-lived I no longer remember what they were called.
My takeaway is that language learning apps are a lot like dating apps. They profit less if people actually learn a language and so can't be too good at their job because they'll bleed users faster than they can gain them - similar to dating apps. It needs to work just well enough that users are tricked into believing it is working but not so well that it actually works for most people.
It seems like the ETA before enshittification begins is about 2~3 years. If you're an early enough adopter you might actually benefit from it but you have to be willing to jump ship and not fall for the engagement/gamification tactics that keep you sticking around after it has stopped providing any value.
I spent way too long 'watering my garden' on Memrise before I looked around and noticed all of the once useful community-providing mnemonics were gone, you couldn't correct bad definitions anymore, it was difficult to actually speak to anyone else in the community (unless you could find them on the forums), and eventually I stopped using it altogether. The community I had signed up for and was a huge part of Memrise's success no longer existed.
"They profit less if people actually learn a language and so can't be too good at their job because they'll bleed users faster than they can gain them - similar to dating apps."
This might be why some of the best language-learning content I have seen is from national broadcasters like YLE in Finland. There, once a foreigner learns the language from that material, they then become a consumer of the broadcaster’s main content.
This is actually a remarkably common failure pattern for a lot of language learning apps. Devs see Anki and think "I'm gonna do it better! I'm gonna build Anki, but for a specific language and make it a web app." ... I've lost count how many of these I've seen over the years!
> It seems like the ETA before enshittification begins is about 2~3 years
It's funny, I actually learned the term "enshittification" specifically from friends of mine who were Memrise users. It's honestly a textbook example of the phenomena.
Negativity aside though, I'm actually pretty optimistic about this space despite all that I've seen so far. I think that there's genuinely room to build great language learning software that people will really benefit from. I'm just really pessimistic about most of the people working in the space. Without trying to exagerrate, I'd estimate that probably less than 10% of people working in the language learning industry are actual language learners (at best, they might've learned English as a kid). When you're not actively, seriously learning a language, you become numb to the problems of people who actually care about becoming fluent and just end up building tinder-esque games to addict people with.
Anki for vocabulary building, Ryan Estrada's comic for learning to read Hangul (https://www.ryanestrada.com/learntoreadkoreanin15minutes/) as it sticks true to its promise. Over 8 years ago I spent 15 minutes learning how to 'read' Hangul. To this day I can still slowly sound things out and, at the least, read people's names. It truly is a fantastic writing system although I do sometimes struggle with which vowel is which that's 100% an issue of only having spent 15 minutes learning.
Unfortunately I can't help much with learning grammar as I never dove into actually learning Korean due to a dislike of how it sounds. There's the "Tae Kim Japanese Grammar"-like approach for a Korean grammar guide at: https://www.howtostudykorean.com/ although I'm not a big fan of how overly simplified (and sometimes wrong due to the simplification) Tae Kim's approach for Japanese was. So I can't attest as to whether How To Study Korean makes the same mistakes or not.
As for writing - Korean is simple enough to read/write that you can simply find any Korean news source and practice writing the sentences as you read them.
> Unfortunately I can't help much with learning grammar as I never dove into actually learning Korean due to a dislike of how it sounds.
Ha, one of my main motivations for wanting to learn Korean is how beautiful it sounds to me. Funny how that goes, diametrically different subjetive perceptions.
It's amazing that something that can look so alien to Western eyes is actually pretty straightforward once you try to learn it. I did the same and learned Hangul so I can at least sound things out and do some basic Internet searches etc.
You can do exactly the same with other scripts, e.g. Japanese hirigana and katakana, which are fairly easy to learn, and also Arabic, which looks difficult, but is definitely learnable in an hour.
I started learning Korean but never really got the far. But, straight after learning Hangul you get into sound mixing (https://www.missellykorean.com/korean-sound-change-rules-pdf...). Trust humans to invent something simple and then make it complicated over time!
Japanese has similar stuff with their u-dropping, but not as complicated as Korean.
It's such a common aspect of languages too that many people don't even realize when they're doing it in their native language!
People get lazy - especially natives who don't get confused because it's how most natives will talk. For example, every word ending in "ing" in your comment could drop the "g" sound when spoken. Plenty of English speakers do that when speakin' and not many people would think anythin' of it until a frustrated person learnin' English asks why nobody is pronouncin' the endin' "g". Droppin', changin', and blendin' sounds is why learning a language by listenin' to natives speakin' is so important instead of crammin' textbooks all day.
Some might consider this a regional thing/accent and I'd argue that it both is and isn't. To the extent I tried to illustrate it would likely get seen as an accent but the occasional droppin' of it is somethin' I've heard across so many different English accents that I'd argue it isn't only an accent thing.
In the US it's mostly associated with a Southern accent and in England it would be the English Midlands like Brummie or Mancunian.
I find far more often that bad UX is the result of someone trying to use a tool for something it wasn't designed for. They might even clob several different tools together in an unholy abomination to get it to do what they actually want instead of having a tool built to do precisely what they want (and once that tool has been built - people will inevitably misuse it to do things other than what it was designed for and then complain about its poor UX for doing those things).