[nerdjon]

This is the problem with all of the recent “AI” crap that has been shoved into our devices.

We have had ML features for years and it provided real benefits but most people did not know or care how it worked, it just did its job in the background without the underlying tech being shoved in your face.

Everything AI though is the opposite, it wants to focus on the technology first and the benefits second. It is actively making a worse UI and often providing little to no benefit.

Most consumers don’t actually care how their tech works, just that it does and gives them benefits.

[dngray]

The real thing i think people forget is that humans actually value time and effort from other humans. AI is often used by people who want to do neither and that's really what it boils down to.

Ask yourself, would you like to receive a christmas or birthday card with a personalized message or something produced you know was 100% produced by AI bot - even better when it has a hallucination in there.

[inigyou]

But that's the rare exception. Almost nobody prefers an artisanal chair over one from IKEA, especially when they see the price tag.

[estaroc]

I think most people would in fact prefer an artisanal chair if not for the price, not just "especially" accounting for price. Not a good comparison here though, because most products are not cheaper to the consumer due to AI - only cheaper (in theory) to the provider.

[ryukoposting]

Bingo. They weren't paying for gmail before, and they still aren't paying for it now that it's more annoying than it used to be.

I suspect that, in many cases, AI features actually make a product more expensive for the operator. Imagine how much of doordash's money you could burn by telling its chatbot that the only way for you to figure out where the driver left your order is to create a todo app in React.

[lbrito]

Everything about that analogy is wrong.

Everyone would prefer a nicer handmade chair (if not by the price difference).

Chairs are not comparable to OPs cards; writing on a card costs nothing (but intent, which seems to be in low stock these days).

Finally, factoring in the real operating cost, ongoing capital costs, and environmental/social externalities, the AI chair in your example would cost something like 1000x a handmade chair.

[partyficial]

> Everyone would prefer

far from true. I (and many others) prefer cookie cutter things, because I know what to expect, and due to quality control of big companies.

e.g. hotels > airbnb, amazon essentials > etsy hand-knit, Lays chips > Ray's chips

[THansenite]

This is so true. My wife loves knitting and frequently gets comments about her items of people asking if they could have her knit something for them. When she tells them that if she tripled the prices of a similar store-bought one, she'd still be making sweatshop wages, they go back to the mass produced version they already have pretty quick.

[HumblyTossed]

Wrong. Most people would definitely _prefer_ the artisanal chair. However, they can't _afford_ the artisanal chair so they _settle_ for the IKEA chair.

[sylens]

There is a distinct difference between a chair and a communication (birthday card, letter, email, whatever) about some personal life event

[aianus]

I bet someone said the same thing about an email or an IM vs a handwritten letter at some point in the past but here we are.

[52-6F-62]

Those are not comparable.

One is the method of recording a message, the other is having something else completely draft a message for you.

The importance of a personal message is not just in the visual appearance or delivery, but that there has to be some emotional loading to even put the effort into drafting one.

With AI, it's a stupid prompt to get it to write trite poesy. It's meaningless and empty at its root. It's discourse with a nullity.

Nobody who values the human connections in their lives wants that. No matter what kind of marketing and fine print gets shoved and manipulated into their lives.

[mattkrause]

And it's still true!

I appreciate a birthday text/email, but a thoughtful email is nicer and an physical card/letter is better still.

[bigstrat2003]

Everyone prefers an artisanal chair over one from IKEA. The only reason they go for the latter is because that's what they can afford, not because they would get that option if all else were equal.

[morgoths_bane]

That one from IKEA was designed by actual people though. If I had a chair designed by robots it probably wouldn’t be as nice, comfortable and evidently, affordable.

[nightski]

Hallmark built a brand on creating generic messages in card.

[throw4847285]

You're supposed to write an additional message inside the card...

[b40d-48b2-979e]

This is an incredible comment, and now I can just imagine the dismay of people receiving a card from the parent poster.

[nightski]

I'm pretty sure you could do this with an AI card as well...

[bluGill]

They have enough different cards that at least you know effort was put into choosing the card. Also effort was put into buying the card at a store, signing it (often/hopefully with a short message), and sending it.

[morgoths_bane]

Perhaps, but you can tell that these cards were made by real people. The art, messages, etc. are all different with varying levels of humor, seriousness, etc. LLMs really seem to converge to specific patterns regardless of the task that are instantly identifiable and low quality. What ends up happening is that the person on the other end thinks “this guy is a moron, he needed a robot to write out a happy birthday card?” Sure getting the hallmark card is mass produced, but they really do hire real artists and writers to make these cards. The robot creates a converged output that is instantly identifiable and has perceived lower quality.

[afavour]

I don’t really want to defend Hallmark too much but I’d argue they provide a means of low effort personalisation. You choose a design that reflects you and your relationship to the recipient. You write a personal message inside (hopefully). The alternative is creating a card from scratch which is a big step up in creativity and time requirement.

[kibwen]

If you don't actually take the time to write something manually inside the card, that's as thoughtless asking an LLM to generate a birthday message to someone.

[devsda]

If you send cards via a third-party subscription service that mails random cards to uploaded contacts in bulk, you'd get marks for setting it up the first time but as the cards pile up and the if the service is detected, it will eventually become a signal that you don't care enough to send them personally.

[threetonesun]

I've seen multiple examples of software with good working ML solutions toss them aside for generalized AI with worse results. The real shift here is an attempt at the "one input for everything" user interface without understanding there's extremely few use cases where that's actually the best interface for users.

[throw7]

They've regressed for a long time and there's no signal to consumers that "AI" is anything that "fixes" or brings back what was working.

The example I always give is when google maps got speech recognition, I could ask it "Hey google, what's the E.T.A." and it would magically respond with how long till I arrive. Somewhere along the line it broke and for years now it doesn't work... the last time I tried my phone actually brought up the web browser and did a web search. smh.

The first thing I did when they forced gemini was I went to look how to disable it. Why? It override the old voice I chose to read calender events in the morning... in fact it would start reading like normal, then that stupid gemini voice would cut in and be entirely unhelpful.

its all enshitification.

[krupan]

Oh my gosh, thank you for writing this so that I know I'm not going insane. I keep thinking there's no way things have gotten worse, like maybe I'm miss remembering? But I was pretty sure I wasn't miss remembering

[throwatdem12311]

“AI” is a buzzword now thanks to the Vulture Capitalists.

The feature should speak for itself. If your feature is good you don’t need to market the underlying technology.

Like, nobody gives a shit about settings being stored in an SQLite database. They don’t care how it’s stored at all.

When my friend shows me his new phone and how crazy it is he can zoom so far into the moon you can see individual rocks - he does not give a single shit that it uses AI. He just uses the gd camera.

When you use AI to build a feature, the fact that it uses AI should not be on the tin. What it actually does and how good it is at it should be. Saying something uses AI is pointless. No matter how much the vulture class wants it, fetch is never going to happen.

[nonethewiser]

Uhhh kind of. What you say is definitely true of some products but it's funny, because the EXACT same criticisms were levied against machine learning.

- "ML is such a buzzword. Everyone is trying to shoe-horn it into their product."

- "Why are they putting 'machine learning' in their hero section? Just do the thing well. ML is an implementation detail."

- "You dont even need ML for this. Simple linear regression would be the better choice."

We are so far beyond the pale. This was a valid criticism ~5 years ago and now we remember it as the golden days.

[nerdjon]

There is a difference though, all of the talk about ML was almost exclusively in the tech circles. Or at most there was a quick reference to "ML" when a feature was announced but it wasn't shoving "ML is doing this THIS" in every UI it could.

Sure we could argue that there were times that ML was likely not really necessary, but it was still largely invisible to the user what the mechanism was.

I think about autocorrect, sentence completion (or just next word recommendation), music recommendations, etc. All of those were clearly ML but the user was not made aware of that at every step of using them and in many cases it being ML was only in technical documents or the original announcement.

Now obviously there are exceptions to this, but it was the exception that shoved ML in your face compared to the current situation around AI.

[vanuatu]

While I agree that AI is more salient, I feel like there was a ton of press about the "Algorithm" especially around social media and content, which is essentially "ML"

[nonethewiser]

> here is a difference though, all of the talk about ML was almost exclusively in the tech circles.

No, not at all. That was a chief complaint. Grandma doesnt give a fuck about machine learning, why are they advertising it?

> I think about autocorrect, sentence completion (or just next word recommendation), music recommendations, etc. All of those were clearly ML but the user was not made aware of that at every step of using them and in many cases it being ML was only in technical documents or the original announcement.

Right. And that's why this isnt an example of the phenomena. Nowhere did I say machine learning was useless.

[Ensorceled]

This is only true on HN. My parents and siblings and cousins and non-technical friends don't even know what the fuck ML or Machine Learning is ... but they all hate AI because they have seen everything AI gets pushed into now sucks and are tired of the AI slop on Facebook and in their Google searches.

[torginus]

I mean I did data science and ML (more the engineering part) back when it was popular, and I kinda echo those sentiments.

ML folks told me their methods would only outperform traditional ones if we could feed the training pipelines an order of magnitude more data than we had. And if a process we're trying to predict only produces X amount of data over a meaningful timeframe, getting 10x, or 100x the amount usually requires 'lateral thinking', which even at the best of times meant either building mass-survelliance into the product or pulling in tangential data, to get marginally better results at best.

What actually worked was understanding what data we had, what it meant, what we missed, and after gathering that right information, even the simplest mathematical tools yielded results.

They still ended feeding the entire thing into 'fancy algorithm X' from scikit-learn, crunched the numbers on those fancy GPUs and got a 5% better result than what the simple methods gave us.

It was still framed as the triumph of these fancy algorithms and expensive data scientists and not people with domain expertise working hard to find and get the right data.

I'm not even salty at this point, I'm just kinda disappointed that a big org has such a hard time doing the right thing for the right reasons.

[ardacinar]

Citation needed. Machine Learning was NOWHERE NEAR as overused as AI in user-facing communication.

The last one is a traditional nerd criticism though, it has been present on HN for the last ~20 years. Kind of ignorable.

[kridsdale3]

The broad public did catch on. They just know there is some invisible force out there named The Algorithm that acts like some fickle god they must appease in order to do well on the internet. Nobody can explain The Algorithm to you, because it isn't like what we learn in school or write in C, it's weights.

[nonethewiser]

You missed my claim:

> Machine Learning was NOWHERE NEAR as overused as AI in user-facing communication.

I never said it was as overused. I said we levied the same criticisms (buzz word, implementation detail, bad fit)

Im surprised people dont remember this. Here are some examples:

- Blind post from 2018 stating machine learning is a buzz word with lots of agreement: https://www.teamblind.com/post/is-machine-learning-bullshit-...

- Tech crunch article from 2020 calling machine learning a buzz word and noting how its advertised prominently: https://techcrunch.com/2020/02/25/when-that-ai-company-isnt-...

- Substack article from 2017 challenging tech companies on if they really need ML: https://makecents.substack.com/p/https-medium-com-medhaa-att...

- Article saying 90% of machine learning is just "linear regression in a trench coat" but people are labeling it ML for marketing purposes: https://www.eigenmagic.com/2022/03/22/linear-regression-is-b...

- More blind posts saying ML is overhyped: https://www.teamblind.com/post/is-ml-a-hype-gxtuchm0, https://www.teamblind.com/post/is-machine-learning-overhyped..., https://www.teamblind.com/post/when-will-this-ai-hype-die-8z...

- hackernews thread for article and comments claiming its overhyped, over-applied etc. https://news.ycombinator.com/item?id=27149532

Note that I am NOT arguing ML is useless. It's not useless. I'm saying people made these same criticisms they make against AI: buzz word, implementation detail, often unnecessary.

It's funny to me that people forget this. I agree the AI buzz is more pervasive. But thats a difference in degree and not kind.

[andrewstuart2]

Not to mention how many features that used to work have been summarily broken with no indication on whether we'll ever get them back, due to the wholesale replacement of previous functions with LLM-driven functions.

I can't even press the "favorite" button for my google photos on my google home device any more. It just says "I don't have access to photos" whether I use the button or voice (both of which obviously used to work).

[yanis_t]

Exactly. It was looking as though Apple understood this. But now they gave in and called it Siri AI.

[nerdjon]

The situation with Apple is what really annoys me about this entire situation, they clearly felt pressure because there was article after article about "Apple falling behind" on AI.

And there is some truth to that given that the features we were supposed to get in iOS 26 did not come out. But it also was just that they were not shoving AI into every single thing.

I still have hope that they will be the company that will (mostly) apply AI in a more meaningful way instead of it just being "AI magic" in everything. There were some genuinely useful things shown at WWDC.

Will have to wait and see though. I was disappointed to see them leaning more into the same branding.

[Forgeties79]

It’s because they don’t know the actual benefits yet and are all hoping they either accidentally stumble across it/one of us finds the billion dollar application for them.