Skip to Content

Artificial Intelligence That Makes Your Smartphone Smarter

Startup Weave.ai is building an alternative to Google Now on Tap that can mine tweets for context and bring up relevant data in other apps on your phone.

Despite their ubiquity, smartphones are still not very helpful at getting you information based on what you’re already doing. For instance, if you get an e-mail from a friend asking if you want to check out a new restaurant in town, you have to leave the app behind and go conduct a Web search to learn more.

An early stage startup called Weave.ai is trying to make this less of a chore by building technology that brings up information that takes into consideration what you’re already doing on your handset.

The London-based company has built a demo of its technology that gives a sense of what it’s building by analyzing the content of tweets and bringing up related information in other apps on the phone (although since it’s still its own separate app, it’s not yet integrated within other apps for easy access).

If what Weave.ai is doing sounds familiar, it’s because it’s not the only one that thinks this kind of contextual search is a good idea. Google showed off a potential solution to this in late May at its annual developer conference: an update to the personal-assistant software, Google Now, coming later this year in the next version of Android. Called Now on Tap,  it aims to bring contextual search to all kinds of smartphone activities—if a friend e-mailed you about a movie, for example, you could press down an Android phone’s home button to pull up an on-screen card with reviews, a link to a film trailer, and other information (see “Google Rolls Out New Automated Helpers”).

Weave.ai cofounder Rodolfo Rosini says the startup expects to roll out a private beta test of an app showing a few different functions in the next few months. The company, which is participating in Techstars’ accelerator in London, plans to let developers add the technology to apps and hopes that it could eventually be added directly to operating software.

Weave.ai cofounder Stéphane Bura says the idea is to build software that lets a search engine focus on what you’re doing at a particular moment with your phone—rather than starting with a blank page that you must enter keywords into.

In one video I saw of Weave.ai in action on an Android handset, the user clicked on a tweet reading, “Saw Age of Ultron. Mixed feelings. Could anything live up to the original?” and then opened up the demo app. The app pulled up a list of results (some related, some not), like a link to the film on IMDB, a link to buy tickets to see it within the app for U.K. movie-theater chain Cineworld, and a Google News story about its second-week box-office ticket sales. A tap on a Wikipedia entry for “Ultron” pulled up a page about the fictional supervillain within that app on the phone, while a tap on the IMDB link led to the listing for the movie in that app.

Bura says Weave.ai works by determining topics related to the words in the tweet, matching them to the apps on the phone that might be most helpful, and querying a database to get a better idea of whether words are, say, a movie title or name. At the same time, it looks up words and phrases from the tweet that it thinks may be interesting on news feeds and places like Wikipedia. Weave.ai uses all these results to search within the apps for relevant information, presenting results to the user as short summaries with a deep link to the information within the app itself—a movie page on IMDB or a restaurant page on Yelp, for instance. Rosini says Weave.ai is trying to figure out not just what was tweeted but also who said it and what your relationship is to that person in order to make the results more relevant.

Beyond just figuring out more ways for Weave.ai to be useful and making its results as relevant as possible, the company may struggle to get developers interested, with Google already announcing its plans for Now on Tap.

Yet Rosini notes that Now on Tap will be available initially for the upcoming version of Android, leaving lots of people with older versions out of the loop. And he thinks many companies—phone makers, social networks, and more—will want to come up with their own version of contextual search, rather than delegating it to Google, in order to do things like control any results that could lead users to buy things (in theory, at least, whoever is in charge of the AI could control affiliate revenue generated by such purchases).

He thinks Weave.ai needs to figure out just a few problems to solve within contextual search, rather than trying to help with lots of things. Because of this, he says, Weave.ai plans to focus on a number of business-type uses, such as analyzing an e-mail about taking a trip to visit a client in L.A. and returning links to book an appropriate flight, rental car, and hotel room within other apps.

“It’s about picking your battles, right?” he says.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.