Amazon may want Alexa to hear and do more. 

The company filed two patent applications detailing ways to make its virtual assistant more useful. To start, Amazon is seeking to patent a way to collect “multi-session context.”

Amazon’s tech stores user data upon initial interactions, including “action data,” or information about user behaviors, “entity data,” or data related to the user, and a profile identifier. The system then holds onto that data for subsequent interactions and uses it to modify responses.

For example, if you ask Alexa to place an Amazon order for a certain product, it may access stored information from previous discussions relating to brands, sizes, or product details. If a user requests that Alexa play “party music,” it may access stored data relating to previously requested songs. Storing previous dialog information in this way may help in “moving the dialog along in a manner that results in a desirable user experience,” Amazon said.

Along with teaching its virtual assistant active listening, the company may want Alexa to multitask with a patent for “multiple skills processing.” As the title of this patent implies, Amazon’s patent aims to help its assistant perform “complex goals” using multiple skills and applications. 

With a user’s permission, data is forwarded between applications to perform different requests, “even if the other action for the other skill was not necessarily requested by the user,” the company said. “Thus the system enables the system to perform multiple actions and the user to initiate multiple skills with fewer user inputs.” 

Amazon’s patents point to a common theme as AI adoption continues to ramp up: Context is key. The more user context an AI model has to work with, the more it can do, said Bob Rogers, PhD, co-founder of BeeKeeperAI and CEO of Oii.ai. “The most annoying thing is an AI recommendation that really doesn’t make sense,” he said.

Obtaining more context allows a system such as a virtual assistant to better connect the dots, understanding the patterns of a user’s day-to-day life. That data ultimately could be used for Amazon’s financial benefit, said Rogers, allowing it to find gaps in user’s lives and make recommendations accordingly. 

“If they are able to create a sort of ongoing series of meaningful threads and context about the people who are using these systems, they’re going to have something very powerful,” he noted.

Plus, giving these tools more insight ultimately makes them more useful, said Rogers. And with competition from the likes of Google (and now Apple) in the consumer AI market, adoption is likely front of mind for Amazon. 

However, the question of privacy versus utility always looms when discussing consumer AI. Though the average consumer struggles to trust AI, these tools are often not explicitly marketed as such. Plus, Rogers said, if the convenience matches or outweighs the data trade-off, consumers aren’t likely to care. “(Amazon) knows about what we watch on TV, they know what we buy, they know when we’re home,” he said. “I don’t think privacy concerns are going to stop people from adopting the tech.” 

“If you ask someone if they’re concerned about data privacy, they’ll say yes, but when it comes to making a decision about clicking ‘yes’ on an end-user license agreement, they don’t really read the details,” Rogers added.

These patents also highlight the growing level of automation that tech firms want, said Rogers. Think of it like a domino effect: While the tech may start by collecting one kind of data or accessing just a few apps, it may be hard to tell where the line is drawn if that convenience grows. “(The patent) feels like they’re thinking about Alexa capabilities, but truly, these things always ended up being ecosystems,” said Rogers. 

Snap wants to let users change their voice. 

The company is seeking to patent a system that uses neural networks for “changing characteristics of vocals.” The patent details a way to change a user’s audio, such as swapping a lower-pitched voice for a higher one, using AI. 

Snap noted that conventional techniques of voice swapping don’t offer a lot of flexibility, and “require a great deal of computation that may be beyond what can be provided by a mobile device.” 

Snap’s patent relies on a convolutional neural network for this process, or one that is particularly good at analyzing visual data. Snap’s system first transforms the audio data into a visual representation of the audio’s “frequencies and intensities,” or essentially a spectrogram.  That image is then fed to the convolutional neural network, which outputs an image with different characteristics such as changes in pitch or timbre. 

To make this system more robust, this AI is trained by transforming audio from having one characteristic to having a different characteristic, such as from a high pitch to a lower pitch, and then reversing the process, such as going from low to high. Finally, a “discriminator” neural network is used to basically grade their work to determine if the AI needs retraining. 

The goal is to provide realistic voice altering capabilities with less computing power and local processing in real time. 

This isn’t Snap’s first interest in AI. Last August, the company launched its own in-app language model, called My AI, though the company received backlash due to its tendency to offer up inappropriate responses to young users and has since added parental restrictions. Snap has also filed several patents for AI-powered inventions, though it is often involved in patents for augmented reality devices or applications. 

But Snap has stiff competition. In May, Snap CEO Evan Spiegel told Bloomberg that the company plans to increase spending on AI and machine learning products to catch up with the competition. The report noted that the company may spend around $1.5 billion annually on AI infrastructure. 

“There was a recognition that we’d fallen behind the curve on the machine learning side, which, to some degree, was reflected in the business performance,” Spiegel told Bloomberg.

Despite its late start, Snap has a lot to gain from AI, said Tejas Dessai, research analyst at Global X ETFs. The company could use AI to strengthen targeting of its digital ads, said Dessai, which is the company’s main revenue stream. “Snap can learn from Meta’s recent success with AI,” he added.

Plus, on the user end, the company could use AI to create more engaging experiences and “significantly strengthen Snap’s standing as a legitimate content platform,” said Dessai, an ambition that this patent seemingly adds to.

“They have a lock on young users, and they’re privy to private messages, in a great position to capitalize on this,” said Dessai. “It’s all about retaining users on the platform and making sure they don’t leave for other apps, and we see plenty of levers at Snap’s disposal to make that happen.” 

The only hitch, however, is user data protection. AI consumes a lot of data, and doesn’t have the best reputation for data privacy. Given that its user base largely consists of minors, Snap may have to tread lightly with which data it uses.

IBM may want to turn you into Inspector Gadget. 

The company filed a patent application for “wearable device recommendations by artificial intelligence.” IBM’s patent details a system that aims to understand the context of a user’s surroundings to recommend what devices or applications should be used at any given time. 

To start, this system collects data about a user’s environment, habits or location through their wearable devices or other “position rendering technologies,” such as cameras or smart home devices. An AI model then  predicts a user’s activity by analyzing that environmental data and comparing it with historical activity data for that location. 

The user is then given a list of recommendations for devices or operations they could use for said activity, “characterized by sensors for capability,” IBM said in the filing. “By employing artificial intelligence the computer wearable devices are matched by capability of their sensor to the user activity.” 

To paint a picture, if a user is at the gym, it may recognize that location, and recommend that the user turn on a heart rate monitor via a smart watch. If a user gets into their car, it may recommend turning on a GPS on their phone. 

Though IBM’s tech is largely applicable to smart watches and phones, it notes that it could be used for a variety of wearable and non-wearable gadgets, including smart glasses and tablets. 

Tech firms seem quite intent on turning AI wearables into must-haves. While Apple and its suite of devices sit at the top of this market, almost every big tech firm is working on increasingly advanced wearables, whether it be smart glasses, watches or headphones. And patent activity from companies like Meta, Microsoft, and Google signals that these devices are only getting smarter. 

It makes sense why these companies want to be with you at every step: The more devices that are strapped to users, the more data they can collect. That kind of data can have several uses, including highly targeted (and highly lucrative) digital advertising and AI training.

This patent takes a similar approach to several other tech firms by discussing wearables through the lens of health and wellness, focusing on things like “bio sensors” for heart rate and blood pressure monitoring. However, the advanced tracking that these devices are capable of could extend far beyond your average fitness tracker. 

Plus, where IBM’s pitch may take it a step beyond is by stringing these capabilities together and making them adaptable, activating certain apps, programs or devices based on the user’s surroundings. If utility is the key to higher rates of adoption, IBM may see a path towards this through accurate and convenient AI-powered recommendations.