4 ms·

This is a serious hype piece. The problem is stated clearly: the smart home industry is expensive, fragmented, and complicated. This isn't news to anyone at all

by Duhck 9y ago

This is a serious hype piece. The problem is stated clearly: the smart home industry is expensive, fragmented, and complicated. This isn't news to anyone at all but there seems to be a lack of a viable solution. Simply connecting every gadget to some variety of interfaces has already been done dozens of times over (e.g. Crestron, control4, smartthings) I'm in the camp that one company should make all the devices versus an open standard. That's the only way to guarantee quality. Candidly, I don't think everything in the modern home will look like Andy's home. Instead our homes will have a few key devices like lights, speakers, locks and HVAC that make sense to interconnect. Once we agree that not every fridge or microwave needs an internet connection, you'll see how ridiculous his vision statement becomes. My 700sqft apartment doesn't benefit from any of this "innovation" and doesnt suffer from the mess of gadgets that don't talk to each other. This is a wealthy persons problem and a tinkerers desires.

CapitalistCartr 9y ago
In a decade every electrical appliance will have a chip linking it to a central controller much as cars work now. Sensors allow the automotive computer to minutely control every detail of the car's functions. In the same way, house functions will become more efficient and automated. I am not at all a fan of this future, but it is coming on fast. Those of us opposed will be like people now stubbornly driving older, less complex autos.
- Duhck 9y ago
  I hear you... commoditization of connected hardware makes this future a real possibility, and yet it hasn't gained consumer traction and it does add real cost to the products. $3-5 of hardware cost + thousands of hours of software adds $30+ of cost to the consumer on most devices. Thats a lot When that changes I'm still not sure it'll happen. There needs to be a consumer demand and today there isn't.
- robolange 9y ago
  Yup. My 1987 Toyota pickup contains no digital computer whatsoever. It's so mechanically simple. I'm pretty sure I could do or have done any major repair (e.g., full transmission replacement, major engine overhaul) for less money than the cost to replace a just-out-of-warranty computer module in a newer vehicle. I realize that computers enable better fuel efficiency, but I just don't see how all that added complexity is worth it in a gasoline-powered vehicle. I'll willingly part with it only when all-electric becomes economically viable.
  - dboreham 9y ago
    If it has a carburetor it isn't mechanically simple.
    - nikofeyn 9y ago
      lol. when i bought my first motorcycle, i explicitly avoided any motorcycle with a carburetor and only looked at fuel injected engines. there's no point in not having FI. and modern cars aren't just mechanically "complex" for the sake of being complex. they are safer in crashes, safer when driving, more comfortable, and are some of the most reliable machines we make. you basically change the oil often and tires every now and then, and that's your maintenance. i drove my last car for 13 years and well over a 100,000 miles (not a lot really), and yet basically zero problems with that maintenance. that's pretty insane reliability for something used everyday.
      - dboreham 9y ago
        You either have no computer but rather complex mechanical things (carburetor, mechanical ignition) or you have computers and quite simple mechanical things (EFI, Electronic Ignition). The computers actually make the car more reliable because there are fewer mechanical things to tweak, tune and fix.
      - lultimouomo 9y ago
        I have nothing against FI, but on my previous motorcycle I could (and did) take apart the carburetor, fix whatever needed to be fixed (old sealings in my case) and put it back, just by reading some stuff on the internet. It is not something very complex. I don't think I could do anything like that with my current fuel injected bike. But then again, I never needed to, because it is objectively more reliable. There is a lot of charm in a computer-free car or bike, but I think most of the aversion people have to electronics is due to the fact that vehicles went through an "uncanny valley" of computerization, where reliability was not substantially better than previous models, yet repairability was substantially worse. I would say we've bridge that already.
- stephengillie 9y ago
  City buses already need to be pulled over and rebooted periodically.
- nine_k 9y ago
  The reason computers are increasingly used in cars is that they make the car cheaper. High-precision, high-durability mechanical parts are expensive. Go shop for a decent motorcycle chain, compare it to the cost of a Raspberry Pi. Removing a complex mechanical part and replacing it with a simpler computer-controller part is a no-brainer.
hinkley 9y ago
Successful product first, open standards second, rich ecosystem third.
paulmd 9y ago
Fully agree that not everything needs internet access, but you listed some useful ones. I don't think this is actually a hardware company at all, to be honest. The actual hardware here is practically trivial, like an ESP8266 that bitbangs the legacy interface. Hopefully some tinkerer would design some PCBs once and just release them as CC, problem solved forever. It's fairly straightforward to send a request to the adapter node and have it do a thing. Again, a hobbyist could probably come up with a reasonably coherent protocol in like an hour or two. There is some small value in a company that can make the act of pairing less painful for consumers, but tinkerers probably won't care. The actual hard part is coming up with software that does useful things with those peripherals/API. He's selling an app, basically.
petra 9y ago
This is a common problem in Silicon Valley: building platforms, instead of focusing on solving valuable problems. But instead let's focus on an important problem many humans have(to a lesser or greater degree) - bad memory. And it's very valuable - employers would be happy for better employees that do less errors, and people would really appreciate better memory. So there's money in this. But if you think about it - the smartwatch could be a great memory aid[1]. And i've seen some research showing smart watch based checklists could help nurses reduce error rates. And another research app that could help people with Alzheimer. But nothing on the market. And zero effort from the big guys to work on that specific problem. And thus we get the smartwatch - a useless gadget that solves tiny insignificant problems. And thus it failed. And it's the same thing with the IOT. Solve the elderly fall detection problem, well. Solve the security problem, well. Find a few more key problems like this and solve them. This will make your solution a must have. Than you could build a good platform and convince others to join(assuming you're not too greedy). And as for the quality problem you mentioned? There are ways to create platforms that guarantee quality. For example, a rust based microcontorller OS, that severly limits what apps can do. [1]Think of having voice activated checklists[1](to reduce errors at your job or home) that are always there. Or after having an important conversation , saying "remember", and the watch remembers. And of course voice activated search. And context dependent search(using something like a high-accuracy location detection) to better narrow search and effectively present contextual info. Or use knowledge from psychology about how our memory works,

by Duhck 9y ago

CapitalistCartr 9y ago

In a decade every electrical appliance will have a chip linking it to a central controller much as cars work now. Sensors allow the automotive computer to minutely control every detail of the car's functions. In the same way, house functions will become more efficient and automated. I am not at all a fan of this future, but it is coming on fast. Those of us opposed will be like people now stubbornly driving older, less complex autos.

Duhck 9y ago
I hear you... commoditization of connected hardware makes this future a real possibility, and yet it hasn't gained consumer traction and it does add real cost to the products. $3-5 of hardware cost + thousands of hours of software adds $30+ of cost to the consumer on most devices. Thats a lot When that changes I'm still not sure it'll happen. There needs to be a consumer demand and today there isn't.
robolange 9y ago
Yup. My 1987 Toyota pickup contains no digital computer whatsoever. It's so mechanically simple. I'm pretty sure I could do or have done any major repair (e.g., full transmission replacement, major engine overhaul) for less money than the cost to replace a just-out-of-warranty computer module in a newer vehicle. I realize that computers enable better fuel efficiency, but I just don't see how all that added complexity is worth it in a gasoline-powered vehicle. I'll willingly part with it only when all-electric becomes economically viable.
- dboreham 9y ago
  If it has a carburetor it isn't mechanically simple.
  - nikofeyn 9y ago
    lol. when i bought my first motorcycle, i explicitly avoided any motorcycle with a carburetor and only looked at fuel injected engines. there's no point in not having FI. and modern cars aren't just mechanically "complex" for the sake of being complex. they are safer in crashes, safer when driving, more comfortable, and are some of the most reliable machines we make. you basically change the oil often and tires every now and then, and that's your maintenance. i drove my last car for 13 years and well over a 100,000 miles (not a lot really), and yet basically zero problems with that maintenance. that's pretty insane reliability for something used everyday.
    - dboreham 9y ago
      You either have no computer but rather complex mechanical things (carburetor, mechanical ignition) or you have computers and quite simple mechanical things (EFI, Electronic Ignition). The computers actually make the car more reliable because there are fewer mechanical things to tweak, tune and fix.
    - lultimouomo 9y ago
      I have nothing against FI, but on my previous motorcycle I could (and did) take apart the carburetor, fix whatever needed to be fixed (old sealings in my case) and put it back, just by reading some stuff on the internet. It is not something very complex. I don't think I could do anything like that with my current fuel injected bike. But then again, I never needed to, because it is objectively more reliable. There is a lot of charm in a computer-free car or bike, but I think most of the aversion people have to electronics is due to the fact that vehicles went through an "uncanny valley" of computerization, where reliability was not substantially better than previous models, yet repairability was substantially worse. I would say we've bridge that already.
stephengillie 9y ago
City buses already need to be pulled over and rebooted periodically.
nine_k 9y ago
The reason computers are increasingly used in cars is that they make the car cheaper. High-precision, high-durability mechanical parts are expensive. Go shop for a decent motorcycle chain, compare it to the cost of a Raspberry Pi. Removing a complex mechanical part and replacing it with a simpler computer-controller part is a no-brainer.

hinkley 9y ago

Successful product first, open standards second, rich ecosystem third.

paulmd 9y ago

Fully agree that not everything needs internet access, but you listed some useful ones. I don't think this is actually a hardware company at all, to be honest. The actual hardware here is practically trivial, like an ESP8266 that bitbangs the legacy interface. Hopefully some tinkerer would design some PCBs once and just release them as CC, problem solved forever. It's fairly straightforward to send a request to the adapter node and have it do a thing. Again, a hobbyist could probably come up with a reasonably coherent protocol in like an hour or two. There is some small value in a company that can make the act of pairing less painful for consumers, but tinkerers probably won't care. The actual hard part is coming up with software that does useful things with those peripherals/API. He's selling an app, basically.

petra 9y ago

This is a common problem in Silicon Valley: building platforms, instead of focusing on solving valuable problems. But instead let's focus on an important problem many humans have(to a lesser or greater degree) - bad memory. And it's very valuable - employers would be happy for better employees that do less errors, and people would really appreciate better memory. So there's money in this. But if you think about it - the smartwatch could be a great memory aid[1]. And i've seen some research showing smart watch based checklists could help nurses reduce error rates. And another research app that could help people with Alzheimer. But nothing on the market. And zero effort from the big guys to work on that specific problem. And thus we get the smartwatch - a useless gadget that solves tiny insignificant problems. And thus it failed. And it's the same thing with the IOT. Solve the elderly fall detection problem, well. Solve the security problem, well. Find a few more key problems like this and solve them. This will make your solution a must have. Than you could build a good platform and convince others to join(assuming you're not too greedy). And as for the quality problem you mentioned? There are ways to create platforms that guarantee quality. For example, a rust based microcontorller OS, that severly limits what apps can do. [1]Think of having voice activated checklists[1](to reduce errors at your job or home) that are always there. Or after having an important conversation , saying "remember", and the watch remembers. And of course voice activated search. And context dependent search(using something like a high-accuracy location detection) to better narrow search and effectively present contextual info. Or use knowledge from psychology about how our memory works,