Forgive my ignorance, but isn't an old school microwave with knobs and button even more accessible, thanks to tacile feedback? Why do they need to put AI in everything?
Anecdotally, the previous tenant of my apartment was blind, and the microwave installed in the kitchen had one of those perfectly flat keypads with no tactile feedback whatsoever. He was able to use it just fine by having some braille labels printed out with a label maker and stuck to the buttons. There was no label on the power setting button, so I assume that, like me, he's just never used it.
He also used these braille labels to know where to turn the knobs on the washing machine, etc.
There are a lot of accessibility improvements I'd think to reach for before AI.
In general, "lot of XYZ improvements I'd think to reach for before AI" should be the default for most problems.
At this rate people will buying SortGPT before trying "sort()".
Though some level of natural language interface does seem more immediately accessible. But maybe just having it read the display back to you would be far easier than involving some kind of AI cloud service.
"Starting: 2 minutes at full power. Press again to cancel"
Bonus points: your house doesn't become instantly inaccessible when the internet doesn't work, or the APIs the device uses are changed, or they "do a Nest". Of course this is a negative point for those hoping to sell you an AI Microwave and an AI Microwave 2 in 2 years time.
Blind person here. This is a lot easier to do if your device requires you to firmly press on the surface to activate the controls. If it doesn't, you will inevitably activate something while trying to feel around for the label. There are ways around this, either a maze of long, thin seals that goes between the buttons and that you can move your fingers over, or a cardboard overlay with somewhat deep holes, but those devices are a lot more annoying.
It's even worse when the device remembers its last state, and going pass the last option cycles back to the first. If this happens, there's no way to reliably set the temperature for example, as you don't know the current value and have no way of setting the device to a predictable state.
Microwave user interfaces are incredibly frustrating.
My parents had one of the earliest commercial microwaves when I was growing up. It had:
- two radio buttons (they were big plastic things, exactly one sticks out and is ready to be pressed at a time) to toggle between cook and defrost.
- a timer knob that had a fin on the front. Turning the fin to 2' o'clock set it to one minute; noon meant "zero / done".
- a paddle switch that opened the door.
- a nice (internal) physical bell that the knob rang when it reached zero.
It was completely accessible for blind people, and was infinitely easier to use than any of the digital models. It took one arm movement to close the door, and push the fin to the right angle.
For the vision impaired, it takes some effort to figure out what the knobs and buttons do; and what state/setting the microwave currently is in (e.g. what's the power level?)
My father-in-law has a special model, he is nearly blind and somewhat deaf - he is 94. He absolutely needs (loud) audio feedback for every operation and countdown for the timer. I can fully appreciate he would not manage with anything less.
Buttons on a microwave are like features on a car (or in a software program). Great for comparative marketing, but do you really need all those features? Do they make it better?
I bought this microwave, currently $74.97 and even my six year old could use it. I'm not blind, but I can't imagine why anybody needs all those buttons. I bought this because it saves me a couple seconds every time I use it. It is a better, simpler design, cheaper, and appeals to my favorite design principle: "what can we remove to make this better?"
The only downside is that the electronics and mechanical parts are not built to last. But that's generally true of the expensive microwaves too, and they have more things that will break rendering the whole thing useless.
https://www.amazon.com/dp/B00BGTNY8O/
Great for comparative marketing, but do you really need all those features? Do they make it better?
Depends what you value. Sounds like you are optimizing for speed of input, so a couple dials are going to be the best option. I like digital for the precise input. I throw a slice of cheese in for 8 seconds because I know that’s how long it takes to get it from cold to to room temperature. It’s a little harder to pick exact, short times on a dial. It is more of a trade off than one type of control is inherently better.
I actually really enjoy the sensor cooking on my microwave. Single press reheat button, defrost by weight works really well, specific food cook modes are spot on (potatoes, popcorn, frozen veggies, chicken). Like, would you not want to just press a button once and have a, evenly cooked chicken breast? Or just put some frozen veggies in a bowl, press a single button, and they're nice and steamy every time?
From the article, it seems there is tactile feedback on the buttons—the author says they aren’t sure which button was the 2 button, implying they could differentiate buttons.
Your solution seems strictly worse. If they’re setting the time with a knob, how do they get 1:35 instead of a guesstimate of the time? How do they change the power level, use the defrost mode, the popcorn preset?
They don’t need to put AI into everything, but having these niceties often help those with disabilities navigate life more easily, and I’m all for it.
I've used many stepped-dial microwaves, and not a single one of them would be good for blind users. The step often changes size after a certain point and steps are often missed. If you're aiming for six and a half minutes, having to do mental arithmetic or keep careful track is not fun ("12 ticks to reach a minute and then 11 ticks of half minutes, and let's hope it actually got them all").
In my direct experience, the best interfaces for blind users are ones that can have tactile/braille labels applied and ones with voice feedback, in that order.
I don’t know - I think for an awful lot of applications, language is the ideal interface.
In 10 years, I want to be able to say “make me some toast and then go feed the cats” to my domestic robot, not sit there twiddling dials to select the toast function and accidentally combining the command with “feed cat” and ending up with “feed me toasted cat”.
Interfaces should be intuitive. Human language is intuitive for all humans who are ever likely to interact with an interface.
What kind of feedback do you get from a grid of buttons? How do I know whether 1 is top left (like a phone) or bottom left (like a keyboard)?
Once the initial setup is out of the way (which I admit is probably a bit of a pain) being able to either talk to it or have it announce what's about to happen seems like a good idea.
> Your solution seems strictly worse. If they’re setting the time with a knob, how do they get 1:35 instead of a guesstimate of the time? How do they change the power level, use the defrost mode, the popcorn preset?
They could use knobs with beveled points and with clear beveled icons/text around the knobs?
I suppose we should ask visual impired person instead of guessing.
Popcorn presets either detect steam once the popcorn is almost done popping (then runs for N more seconds), or actually use a microphone to detect the time between pops (I think the latter is more accurate). If done well, it can work well. Apparently it’s often not done well because the bags of microwave popcorn almost always say “don’t use popcorn setting!”
The problematic popcorn settings are the ones that don't have any such detection and instead ask you to enter a weight and they choose an amount of time based on that. It's very inaccurate and causes either many unpopped kernels or some burning.
Right I'm wondering if I'm missing something too, although some parts were making sense like opening the box and the turntable, some parts read like workarounds to the situation? I hope someone better informed can chime in
Also, LG (of all the appliances we put in our house, LG has, by far, the shittiest software -- they are somehow worse than samsung) is going to move to a mode where you need to have the wifi connected so it can confirm that you've paid for your monthly appliance subscription, and so that the ads stay up to date. LG projects a 50% revenue increase from this (they are not planning to subsidize the purchase prices of their stuff; this is pure double-dipping).
I think they are delusional, but, assuming it works, this means the subscription/ad revenue would add up to ~ 50% the price of the appliance on average.
I guess people will just starve to death or something when their wifi routers go offline, but that's a small price to pay for an integer factor increase of LG's profit margins.
