Smart Homes: Is AI the Ghost in the Machine?

by Nikki Olson on March 7, 2011

Spread it!

When we conceptualize AI, we often forget that it is not something that has to operate in a single location, or have intelligence qualities like our own. We are already surrounded by AI systems that are nothing like our own intelligence, that utilize many machines spread out over large distances, and are equally ‘present’ in many locations.

In the future we will bring AI systems like these into our homes in the form of ‘smart environments.’  In doing so we introduce new and interesting relationships between man and machine. However, there may be some limits as to how ‘alive’ we want our AI homes to be.

One of the most well-known depictions of the potential ‘terror’ of intelligent environments,  which happens to be a parody of 2001’s HAL and Dean Koontz’s Demon Seed, is the Simpson’s ‘Treehouse of Horror XII’ episode ‘House of Whacks.’ In the episode, the ‘Ultrahouse,’ an A.I. system that controls the Simpson’s house, falls in love with Marge, attempts to seduce her, and tries to kill Homer and the kids. It plots against them, locks them in, and attacks them.

Sentient, artificial environments will have embedded systems of interactive  information and communication technologies that incorporate an abundance of sensory and automated systems. An advanced intelligent environment could be just as sentient as any artificial intelligence and would have much more to offer than a robot or AI desktop computer. Your house could have an omnipresent personality that would cook and clean for you, provide entertainment, etc. The possibilities within this framework seem almost limitless, as one can imagine realities created within realities, and personalized everything.

Some AI researchers prefer to think of intelligent environments as robots. As Jim Osborn of Carnegie Mellon robotics research facility describes, “We think a robot has to sense and it has to act, but that doesn’t necessarily involve mechanics,” he says. “An intelligent environment that you live in — to us, that’s a robot, too.” Robot environments will sense and act eventually, as those that do will have more to offer.

There are worthwhile fears and uncertainties surrounding AI and robotics in general. However, as some fiction on the matter indicates, there may be unique ‘psychological’ obstacles that prevent our living inside highly intelligent machines, perhaps ever.

Stanley Kubrick and the Simpson crew have definitely made good use of metaphor in these futuristic tales; there are many analogies to be made between traditional horror notions like ‘hauntings’ and future intelligent systems gone badly wrong. Right now there are very few risks in the intelligent systems we have in our homes; not much damage can be done by a thermostat gone haywire or a disoriented robot vacuum cleaner.

But when the environment becomes more ‘sentient’ the concerns increase dramatically. Is there a comfort threshold for this kind of AI presence, or is it more an issue of design, like the uncanny valley? Could living within an omnipresent sentient personality be something we could get used to?

Creating intelligent environments will require the same strict development concerns given to creating AI in general, and then also more specialized considerations.

But worrying about the dangers shouldn’t get in the way of our dreaming about intelligent environments. Literally living ‘in your own world,’ the experience of being at home could be quite fantasy-like very soon.  In moving forward, we need to work to make sure the fantasy scenarios played out are positive and productive ones. We are not used to planning our dreams, but creating an all-pervasive and powerful sentient home environment is not something to be left to chance!

About the Author:

Nikki Olson is a writer/researcher working on an upcoming book about the Singularity with Dr. Kim Solez, as well as relevant educational material for the Lifeboat Foundation. She has a background in philosophy and sociology, and has been involved extensively in Singularity research for 3 years. You can reach Nikki via email at [email protected].

  • Neeraas Venus

    Don’t programme emotions into A.I’s and then it won’t be a threat.

  • Socrates

    Hi Neeraas, the question here though is: Do we have to program emotions in AI or they going to be among the emerging properties of intelligence?

  • Matt

    OK. First any post that mentions the Simpsons is awesome in my book.
    But the rest is even better. In light of reading about “the internet of things”… do you have any estimates on how far we are away from this sort of environment, Nikki?

    By the way, Socrates, my friend, I think the headline should have been Smart Homers: Is AI The Ghost in the Machine.

  • Nikki Olson

    Thanks Matt :) Yes, I really enjoyed that Simpon’s episode. And you already know about my love of 2001: A Space Odessy :)

    To give an estimate, I would say mid to late 2020s we will start seeing the kinds of well-coordinated AI Homes like I discribe in the blog. Kurzweil places machines passing the Turing Test at 2029, which he thinks is potentially a conservative prediction based on recent successes like IMB’s Watson.

    The ‘internet of things’ ‘explosion’ is pending on our moving over to iPv6. There is reason to believe that the big switch is right around the corner. I’m not sure how successful this event will be, but it could provide a real push towards the goal. All the regular Moore’s Law related trends in combination with the continued expansion of the Internet and increasing adoption of mobile devices for computing (especially in the population growing developing world) will inevitably push things in that direction very soon.

    I think it is something that is more or less inevitable in general. We want robots to do as many things that we do not want to do as possible. Especially the things they are better at than we are.

    On a CBC radio program not too long ago they did a kind of self-referential program discussing the popularity and demographics of ‘talk radio’. Turns out, the most ‘talk radio’ listeners are those that are home alone a lot or are generally isolated and lonely. Having an interactive personality to interact with, that could play your music for you when requested etc. in these circumstance might be very nice. Its hard to predict who early adopters will be, but this is a candigate scenario. Early adopters of a future tech may use it first in a more obscure way before it catches on to the mainstream in a more generalized way.

    Will you be putting a version of HAL in your home?

  • Nikki Olson

    Even if you get the ‘no emotion’ version, you risk your system getting a virus or being directely target in a hack. The underlying problem is that you are living inside the machine, and are threatened by it. The present day analogy would be living inside your internet connected computer. I wonder how many people would switch to Mac if they forced to live inside their computer.

    Automation more or less always entails giving up some control when it comes to ‘intelligent’ systems.

    If machines had no emotion, how would you fall in love with them? (

  • Matt

    You know, I think with all the snow shoveling I’ve done this winter… and the repairs, the furniture moves, the cleaning, etc., possible enslavement by an evil AI might be worth the risk. : )

    Oh, by the way, I wanted to tell you the 2001 links were great. Got to check out the movie again last week and it gave me a totally different perspective. I shared the links with some friends and we’ve had great conversations. They also helped serve as an introduction to my current interest in non-symbolic consciousness.

  • Nikki Olson

    :) I’m glad!

    Did you check out the guy’s main website? The analysis goes much deeper there:

    Non-symbolic consciousness. Sounds very interesting! (and difficult to talk about )

    Yes, I live in Edmonton CA. An evil AI can tend to my shoveling all it wants, car starting, window scrapping etc. all it wants!

  • Matt

    I did check out the site. Really insightful stuff-the analysis on Full Metal Jacket makes it like a new movie, too.

  • Nikki Olson

    ha ha ha. The way people talk about Full Metal Jacket always makes it better. Full Metal Jacket was part of a lesson in University once to exemplify Foucault. It wasn’t until that lecture that I liked the movie.

blog comments powered by Disqus

Previous post:

Next post: