Amazon’s Alexa is one of the best-known digital voice assistants. Amazon has managed to perfect it to a level in which Alexa can flawlessly perform some tasks, allowing the user to multitask with a virtual aid. However, this does not mean that there are no problems with Alexa. Today we will be talking about some hacks that could become malicious if they landed in the wrong hands. They mimic some of Alexa’s commands, and they can even record you without you knowing it.
Moreover, we will also talk about how Google has raised the stakes after its i/o event with some of the new features added to Google Assistant and how do these new features compare with what Alexa already has managed to perfect.
New voices for Alexa
Amazon is now finally allowing coders to play with eight new, different voices for their skills. This is all part of the company’s Polly service, which most developers use to create texts and later on convert them to voices which Alexa can then use. The developers accepted to test these new voices will be allowed to then use them for free. However, do not expect this to last since after the preview is done and the voices will become official this perk may be revoked.
SO what can you do with these new voices? The easiest way to integrate them into a program that Alexa could then reproduce is if the developer decides to create an interactive story. If just one person reads the story then people, especially children, may have a hard time telling which is which. However, if each character has a different voice (selected from those available) then the story has more depth and children will love it even more. This does not just apply to child stories; it could also be done with audiobooks. There are a host of other options as well; developers only have to be creative.
Alexa to be taken advantage of through silent spy hacks
A group of researchers from Indiana University, The University of Virginia, Bloomington and the Chinese Academy of Sciences have decided to team up so that they could determine if Alexa could be taken advantage of. As with any piece of software, there are at times some loopholes that hackers can make a profit from.
The hack has been named ‘voice squatting,’ and we believe the term is quite relevant. The hack works because it uses what we now call “skill markets.” These are third-party services which range from bank apps to entertainment apps. You can open these apps with certain commands. The team of researchers discovered that you could create some “malicious skills” which would sound similar to the regular command. But what is the problem with that?
Say you voice out the command that opens your banking app, but the malicious skill is activated. If you then start speaking about your banking details hackers now have complete access to your personal information and can use it as they see fit. You see the problem now? We wouldn’t want that, and I bet you wouldn’t either.
Another secret hack has been dubbed the “reprompt” function. This allows the skill to keep running if it does not receive a response. However, this can also be taken advantage of. Researchers created a long audio file with nothing on it meant to play while the actual mic records the sounds around it. On Alexa, this feature was able to run for 102 seconds and 264 seconds on Google.
Google Assistant vs. Alexa
Since Google I/O is done we could not help ourselves to make a comparison between Alexa and the new features that have been announced for Google Assistant.
With Google Assistant you will now get six different voice options, apart from the original female and male one. Alexa only has the female option thus far, but her accent can be changed so that she sounds Canadian, Indian or British.
Both Google and Alexa now have new features that will allow the conversation with the AI to be smoother. Instead of saying “Hey Google” every time you want to ask something, Google Assistant now has a new feature which allows you to ask to follow up questions. It can understand that what you are asking is a follow-up question so it will give you the answers you need. Alexa has a similar feature available which lets the user ask a follow-up question a couple of seconds after he or she made the initial request.
At I/O Google decides to announce that Google Assistant is now compatible with over 5,000 smart home devices. However, Alexa is still leading the game since it is compatible with 12,000 smart home devices.
Perhaps the most exciting feature that Google Assistant now has is that it can make phone calls for you. This was also debuted during I/O, where developers called a hair salon. The AI’s voices sounded incredibly human-like, and it even let out small “hmms” and “uhms” during the conversation, even adding pauses during the conversation. If Google Assistant can already do all of that so far and sound almost like a human then what is going to be added to the AI next?
So far Alexa cannot make phone calls for you, but we expect Amazon to come up with a similar feature in the next couple of months so that it could meet its competitor on an even playing field.