Hacker News new | ask | show | jobs
by cptnbob 3936 days ago
100% this. I haven't found a genuine IoT device that is useful. They always remind me of this scene from The Fifth Element:

https://youtu.be/SnzzWGcdMqY

Literally ever device has dubious utility and security so far.

4 comments

I think it is the consumer IoT space which struggles for usefulness. In industry, IoT allows agriculture to use less water and get more yield, as well as monitoring civil infrastructure (roads, watermains, bridges, etc).
That's just embedded systems that have always been there though isn't it? Someone just stuck a marketing label on it.
"Big data" things used to be data warehousing. "Cloud based" used to be "web based". Maybe I a too old and cynical, but I rarely see genuinely new things, just rehashes of old things.
Yes. I think you've just been through a couple of cycles now like myself. Cynicism comes from watching the last disaster unfold :)
But embedded systems (as far as I know) weren't by default internet connected. They weren't easily merged with other APIs (again this may be my limited knowledge) to provide a more holistic vision of a system. They weren't easily accessible to be consumed as valuable data in real-time on phones.

Yes, an embedded system could get the soil conditions on a farm, but what did it do with that data? How was it made available to an end user.

So, I don't disagree that embedded systems have been in existence for a long time, but I think what we're seeing is the equivalent of moving from Compuserve to the WWW, and I think at the time, many people would have said that all the world-wide web was doing was making the information available through Compuserve's forums and networks distributed. But in reality, it was the early stages of a massive sea-change in how the technology is used.

Consumer IoT in its current form is mostly a product of the O'Reilly Hype Machine - the machine that gave the world Web 2.0.

I get some use out of remote heating control. My hours are variable, so a dumb timer isn't ideal.

Lighting and the rest I have no interest in.

Useful consumer IoT is going to have to wait for useful consumer robotics, and that's 10-20 years away.

Consumer IoT is too expensive, not useless. Everything it can't do is a metric of the fact that computing on that scale actually isn't cheap enough.

Light control for example: people don't want to spend the $thousands it would cost to make it practical (i.e. every light in your house control) when the most common use-case it would solve is leaving your house and going "oh, I left the lights on - but click!". That's worth maybe $10 a light to me, not $70. Which in turn implies the wi-fi, compute and power has to cost cents, not dollars in the product.

The associated problem is the blown-up featuritis required to somehow make up for the price.

I would be extremely happy to shed out 70$ for a way to turn off the lights when I forget them (cynically, I'd say especially if it lasts long enough to save me 70$ in electricity bills, but I'm sure a reasonably convincing argument can be made about the environment, additional comfort and so on).

The problem is that virtually every available offering right now attempts to do a lot more than that to justify the 70$ price tag.

It also does it disastrously bad, and without any meaningful integration behind a single manufacturer's product range (there are few things more hilarious than looking at a bunch of Z-Wave or Zigbee devices from different manufacturers trying to talk to each other).

Instead of giving me a simple on-off gizmo, I get a crawling smartphone app that needs a frickin' cloud infrastructure behind it and pushes HTTP notifications to something that could literally be made out of a dozen transistors if it didn't need to talk HTTP (or at least out of a cheap-ass microcontroller with a dumb hardware PHY, but because IPv4 won't fucking die already we either need to make them talk HTTP or have people toy with their routers). When I get downstairs and notice the lights are still on, it takes me less time to find my keys, run back upstairs, fix me a glass of Old Fashioned, gulp it, go back downstairs, notice I forgot why I went up in the first place, so go back up again and turn off the light, than it takes me to open the fucking application, wait for the authentication messages to go back and forth between me, the lights that are literally five meters away from me and some datacenter in frickin' Texas, and turn off the light after I finally get to the relevant stupid dashboard that seemingly shows every possible control in the world except for the on/off button.

All because -- if all it did was shown an on/off button -- it would seem way too expensive.

It's bitterly impossible not to remark that everyone seems to try putting out the fire with the sword. The whole tech stack is currently way too complex and burdening to manufacture these things cheaply. The logic thing to do would be to make it leaner. Yet instead, everyone seems to be trying to get chips to talk JavaScript now.

Well, all lights are controlled by a single breaker on most homes I know about, so killing them all would be simple, with to need for clunky per-bulb wireless devices.

The problem is that most IoT devices seem to be developed with "oooh, future!" in mind instead of more basic practicality.

Aren't you using the Web 2.0 right now though?
What, for a bulletin board?

Kids nowadays...

I was really just teasing the OP. But when you think about it, bulletin boards, being almost totally UGC are kind of the proto-Web 2.0.
https://en.wikipedia.org/wiki/Web_2.0

I can't wait for a similar 'cloud computing' retrospective, aka using 'other peoples computers to compute'.

No. There are no rounded corners here.
Bathroom scales that log your weight to your health app automatically.

Lights that dim when the movie starts.

A front door that unlocks when you approach.

None of these are necessary, obviously, but they offer little conveniences. An iKettle is pushing it a little, but honestly, I'd quite like to receive a tap on my wrist when the kettle has finished or my toast has popped up.

There's plenty of utility to be had from IoT devices (although you're right about security being a concern) however, like remote controls and robotic vacuums, initially they are being dismissed as tools for the lazy. Soon I suspect they'll be part of everyone's lives.

All of these things sound like solutions that will take me a lot longer to set up than they will save me in time. A few years ago I messed about with "smart" lighting, and it turns out that all of it just unimaginably annoying. All of these "easy" devices are going to have horrific security, some day there's going to be articles about a doorknob botnet or exploits running on shovels.
Relevant XKCD: https://xkcd.com/1205/
With this comment, that multi-functional Chinese army shovel becomes so much more sinister.
> A front door that unlocks when you approach.

Mmmh, what could possibly go wrong with that?

I do not deny that there might be some useful applications for "IoT", but what I have heard / read about so far sounds mostly like solutions desperately looking for problems so everyone can ride the hype train.

Given the numerous security issues we have seen on the Internet so far, I have serious doubts about hooking up everyday objects or critical infrastructure to the Internet just for the sake of a little convenience.

> Mmmh, what could possibly go wrong with that?

Honestly? Not much, really. As long as there's a backup in case you lose your phone (some have a regular key lock, some require a passcode) then you'll be fine. Thieves aren't going to hack your smart lock, they'll just crowbar the door or throw a brick through a window.

Critical infrastructure aside, hooking up every day objects like scales and other appliances (I'm looking forward to notifications such as "Your washing is ready" or "Your pizza might be burning") is potentially really useful. There'll be rough spots at first - configuring these devices right now isn't great - but hopefully systems like Homekit are going to help with the setup and administration of these devices.

> Lights that dim when the movie starts.

How does it know you want the lights dimmed? What if someone in the room is reading? This would be pretty high level AI.

>A front door that unlocks when you approach.

It has to know if you'll enter the house. What if you don't?

It could lock it again, but then what if someone else left the house assuming it was unlocked. Now they could be locked out.

These are hard problems to solve and more AI related.

Exactly. One of the best things Joi Ito has done at the MIT Media Lab has been to shift the culture from "demo or die" to "deploy or die," because in that demo moment we are susceptible to the narrow and novel circumstances of the idea as performance.
I suppose a camera in a fridge would be handy, so you could have a peek at whether you were out of milk.

I believe there was a similar product, but you had to swipe product barcodes, to maintain a database. But I'm not the kind of person who keeps a database of the contents of my fridge, which might be why I'm the kind of person who forgets to buy milk.

My heuristic for coming up with that idea - when's the last time I called home, and ask someone to perform a simple task. Pet food dispensers might be another one. I guess a remote kill switch for irons, ovens, etc could be useful.

And then someone prevents you from running said oven until you pay them a ransom.

(Or just starts turning on peoples irons at random, if it can turn them on as well)

I personally think the security implications outweigh the potential benefits for most IoT devices. It's too bad, because the potential benefits are generally nice.

(And the other side is: I have no desire to "buy" a device that can have features remotely removed. It's why I ditched Windows, it's why I won't ever buy a Tesla or an iPhone, it's why I don't have any recent game consoles. And most IoT devices fall into this category. Among other reasons, it creates a perverse incentive for a company to break (intentionally or unintentionally) older models (or allow them to fall into disrepair) so that people will upgrade to this year's model.)