AI (artificial intelligence) is all the rage these days. Hardly surprising, when you consider that natural intelligence seems to get even rarer as the years march swiftly onwards.
People are busier than ever before in human history, with their seventeen job roles, five smartphones, and shrinking personalities. “I’m so busy,” they cry out plaintively, seeking someone to talk to as their human friends have long since vanished—they too are ‘too busy.’ True story.
Luckily, some of the world’s largest tech companies have recognized this trend and spent billions on deep learning to enhance voice recognition capabilities and help these busy people keep from talking to one another. Enter the world of the intelligent digital assistant, which allows all those busy people to once again have conversations, albeit with their smartphones or other enabled devices.
What about the rest of us? Are there options for all platforms? What are the pros and cons?
Apple has Siri, Microsoft has Cortana and Google has Google Assistant embedded in their operating systems, apps and browsers. Amazon, a retailer, has Alexa, and Baidu, the Chinese search engine giant, has Duer (a faux English word, but in Chinese it’s du mi, where du is short for Baidu and mi is short for mi shu, or secretary).
From Portable Assistants to Standalone Devices
It was only a matter of time before these intelligent personal assistants moved from smartphones and hit our homes and offices, incorporated into standalone devices known as smart speakers. Amazon’s Alexa was the first to hit the mainstream with its Echo smart speaker in 2014. It had the facility to add apps or skills and later, to connect with home automation devices. All companies involved in the digital personal assistant space encourage innovation by allowing hardware and software developers to integrate with their devices. Apart from Apple, of course, as Siri is only available on Apple hardware, like their new smart speaker, HomePod.
For many, the smart speaker will become the center of their smart home setup, eliminating the need for smart hubs that were earlier deemed the best route forward.
But given the wide array of devices available in the smart or automation markets for home and business, is a smart speaker useful in the home?
Let’s ask Sorin Mustaca, CSSLP, Security+, Project+, an independent IT security consultant.
How do you feel about the smart home in terms of data/network security and actual premises security?
“I am a big fan of the IoT, because it has the chance of improving several aspects of our busy lives. If such a device is making our lives easier and not more complex, I am willing to give it a try. Unfortunately, things are very far away still from that moment. I am a big fan of smart homes, but I am not a fan of the security issues they create. I think that the entire concept is in its infancy and there will be some years before we can really communicate, and they can understand what we want. And a few more until we can trust them with our homes, private information, and to let them decide on some things on our behalf,” said Mustaca.
Would you rely on a smart device for critical aspects of your home?
“I rely on an AI assistant to play music, turn on lights, stream some content. But no more than that,” said Mustaca.
Given a choice what device would you select, a device from Google or Amazon? Why?
“I have selected Google. I do not think that Amazon is able to make hardware and good software for it. I have tested almost all hardware devices produced (tablets, e-book readers) by Amazon in the last few years and have returned them all. In addition, I can’t think of a digital assistant that tries to sell me goods from Amazon,” said Mustaca.
What is the problem with these devices, regardless of manufacturer?
“These assistants, no matter who produces them, are far from what you see in the advertising. Those features you see are probably the few things they can do well. We are still in the phase where we must learn what to tell them and how to tell them. If you can’t communicate naturally to explain what you want the assistant to do, you will always have errors,” said Mustaca.
Any prediction for the future of smart speakers – interactive homes or offices?
“I see a bright future in the next 3-5 years, if the AI behind them gets seriously improved. I think that when we can interact with them without a list of keywords, we will have reached the moment where universal usage is possible,” said Mustaca.
Valid observations, certainly, as all companies freely admit that it’s early days for voice recognition and it’s no coincidence that even Dragon dictation software (widely considered the best available) involves a training period for the user’s accent, speech patterns and vocabulary.
Speech recognition is lampooned effectively on the Big Bang Theory as a ‘flawed technology’, Siri cannot handle Raj’s surname or in a later episode, Kripke’s Fudd-like queries. Wabbits? Where?
If You Must Buy, Buyer Beware
Google and Amazon are not the only players in town, as indicated earlier, but they are the major players. Google is arguably leading the AI charge in voice and image recognition, has the technical resources in-house to handle deep learning that ties in with a vast array of products and services. Amazon, regardless of size, is a retailer and has a different focus than Google.
If you want a smart speaker and you must buy one as soon as possible, then consider the following. I have listed each in order of importance, at least as far as I’m concerned.
- Is security a key consideration when selecting a smart speaker? If so, stick to well-known brands only and consider a dedicated LAN for IoT devices.
- A smart speaker should have good audio i.e. it should sound good.
- It should not stain wooden surfaces – Apple, take note.
- It should be compatible with your desired intelligent digital assistant, whether Alexa, Cortana or other.
- It should be compatible with the smart devices and sensors you plan to add to your home or office.
- Are you restricted to specific services or can you access all, including those that compete with the product manufacturer? This means your desired services may be key to device selection.
- You realize that the ‘speaker’ aspect of this setup involves constant microphone monitoring in your home and office, with or without a wake word, it is constantly listening. If this is okay, then you are also confident that device providers will manage your audio data securely and only provide it to law enforcement when you annoy them in some way.
- Any smart devices added to your home or office should be secure by design. If not, don’t add them. Pay special attention to home security products.
- Feel free to check all surveys but Google speakers seem to be the smartest – recognizes pronouns for one thing.
In conclusion, a smart speaker is certainly an innovation, but the security dangers are obvious as new devices are added. For the lonely, lazy and those who crave the sci-fi appeal of a talking home or office, interaction with a digital speaker is enjoyable and closing the drapes with a voice command is a true demonstration of power.
In my case, I am holding off on a decision and will not buy a smart speaker, at least until a few standards are finalized, such as those in IoT security and low-power wireless standards. In the meantime, my butt leaves the couch to dim the lights and remotes control my home theater system. I’m okay with that
Once my intelligent digital assistant can detect an incoming hack, warn me and remotely explode the hacker’s bionic implants, I might consider allowing AI to monitor my home through a wide variety of sensors. Until then, all surveillance is actively discouraged. Enter my home at your peril. I’m armed to the teeth… “Hey Alexa, Siri and the rest of the gang, you listening?”