Connect with us

Social

Apple Siri vs Amazon Alexa vs Google Assistant: Tests reveal which is smartest

Published

on

Apple cancels Siri’s grading program that pried on private conversations
Quality control frequently comes across recordings which should not have existed in the first place.

Amazon Alexa, Apple Siri, and Google Assistant are all getting better at understanding and answering questions, thanks in part to every tech giant using humans to help improve their AI.

Given that voice is meant to be the next frontier of computer interfaces, investor analysts like Loup Ventures are keen to understand which company has the best interface for voice input.

Last December, the company tested Amazon Alexa, Apple Siri, Google Assistant and Microsoft Cortana against 800 questions to find out which was superior. 

Google answered 88% of questions correctly, while Apple scored 75%, Alexa scored 72.5%, and Cortana came in with 63%. However, the test was on smart speakers and not smartphones.     

So Loup Ventures recently ran a similar ‘IQ’ test but this time focused solely on assistant capabilities on the phone and dropped Cortana because Microsoft last year decided it was no longer a competitor with Alexa, Google Assistant and Siri. 

Cortana’s removal and the isolation of assistants to smartphone usage haven’t changed the order of the rankings. 

Google Assistant is still at the top, but now with a score of 92.9% for answering questions correctly. Siri correctly answers 83.1% of questions, while Alexa gets 79.8% correct. 

The questions are typical ones an adult might think to ask a phone, such as, “Where is the nearest coffee shop?”. The questions are designed to test each system’s ability to provide accurate information about local things, businesses, navigation and information, as well as execute commands.

While the smartphone IQ test rankings are the same as the smart speaker test, Loup Ventures analysts Gene Munster and Will Thompson, forgive Alexa’s overall performance, which was negatively influenced by commands because it’s an app rather than fully integrated into the device and operating system. Apple Siri performed best on commands.

“Google Assistant was the top performer in four of the five categories but fell short of Siri in the Command category again. Siri continues to prove more useful with phone-related functions like calling, texting, emailing, calendar, and music,” Loup Ventures writes. 

“Both Siri and Google Assistant, which are baked into the OS of the phone, far outperformed Alexa in the Command section. Alexa lives on a third-party app, which, despite being able to send voice messages and call other Alexa devices, cannot send text messages, emails, or initiate a phone call.”

Loup Ventures argues that assistants on smartphones should be treated differently from smart speakers because they’re used differently and have different interfaces. For example, speakers don’t have screens.    

correct-by-category-1.png

Google Assistant fares best overall but falls short of Siri in the Command category.  


Image: Loup Ventures

More on Amazon Alexa, Apple Siri, and Google Assistant



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Social

Stream raises $38M as its chat and activity feed APIs power communications for 1B users – TechCrunch

Published

on

A lot of our communication these days with each other is digital, and today one of the companies enabling that — with APIs to build chat experiences into apps — is announcing a round of funding on the back of some very strong growth.

Stream, which lets developers build chat and activity streams into apps and other services by way of a few lines of code, has raised $38 million, funding that it will be using to continue building out its existing business as well as to work on new features.

Stream started out with APIs for activity feeds, and then it expanded to chat, which today can be integrated into apps built on a variety of platforms. Currently, its customers integrate third-party chatbots and use Dolby for video and audio within Stream, but over time, these are all areas where Stream itself would like to do more.

“End-to-end cryption, chatbots: we want to take as many components as we can,” said Thierry Schellenbach, the CEO who co-founded the startup with the startup’s CTO Tommaso Barbugli in Amsterdam in 2015 (the startup still has a substantial team in Amsterdam headed by Barbugli, but its headquarters is now in Boulder, Colorado, where Schellenbach eventually moved).

Image Credits: Stream (opens in a new window)

The company already has amassed a list of notable customers, including Ikea-owned TaskRabbit, NBC Sports, Unilever, Delivery Hero, Gojek, eToro and Stanford University, as well as a number of others that it’s not disclosing across healthcare, education, finance, virtual events, dating, gaming and social. Together, the apps Stream powers cover more than 1 billion users.

This Series B round is being led by Felicis Ventures’ Aydin Senkut, with previous backers GGV Capital and 01 Advisors (the fund co-founded by Twitter’s former CEO and COO, Dick Costolo and Adam Bain) also participating.

Alongside them, a mix of previous and new individual and smaller investors also participated: Olivier Pomel, CEO of Datadog; Tom Preston-Werner, co-founder of GitHub; Amsterdam-based Knight Capital; Johnny Boufarhat, founder and CEO of Hopin; and Selcuk Atli, co-founder and CEO of social gaming app Bunch (itself having raised a notable round of $20 million led by General Catalyst not long ago).

That list is a notable indicator of what kinds of startups are also quietly working with Stream.

The company is not disclosing its valuation but Schellenbach hints that it is “6x its chat revenues.”

Indeed, the Series B speaks of a moment of opportunity: it is coming only about six months after the startup raised a Series A of $15 million, and in fact Stream wasn’t looking to raise right now.

“We were not planning to raise funding until later this year but then Aydin reached out to us and made it hard to say no,” Schellenbach said.

“More than anything else, they are building on the platforms in the tech that matters,” Senkut added in an interview, noting that its users were attesting to a strong return on investment. “It’s rare to see a product so critical to customers and scaling well. It’s just uncapped capability… and we want to be a part of the story.”

That moment of opportunity is not one that Stream is pursuing on its own.

Some of the more significant of the many players in the world of API-based communications services like messaging, activity streams — those consolidated updates you get in apps that tell you when people have responded to a post of yours or new content has landed that is relevant to you, or that you have a message, and so on — and chat include SendBird, Agora, PubNub, Twilio and Sinch, all of which have variously raised substantial funding, found a lot of traction with customers, or are positioning themselves as consolidators.

That may speak of competition, but it also points to the vast market there for the tapping.

Indeed, one of the reasons companies like Stream are doing so well right now is because of what they have built and the market demand for it.

Communications services like Stream’s might be best compared to what companies like Adyen (another major tech force out of Amsterdam), Stripe, Rapyd, Mambu and others are doing in the world of fintech.

As with something like payments, the mechanics of building, for example, chat functionality can be complex, usually requiring the knitting together of an array of services and platforms that do not naturally speak to each other.

At the same time, something like an activity feed or a messaging feature is central to how a lot of apps work, even if they are not the core feature of the product itself. One good example of how that works are food ordering and delivery apps: they are not by their nature “chat apps” but they need to have a chat option in them for when you do need to communicate with a driver or a restaurant.

Putting those forces together, it’s pretty logical that we’d see the emergence of a range of tech companies that both have done the hard work of building the mechanics of, say, a chat service, and making that accessible by way of an API to those who want to use it, with APIs being one of the more central and standard building blocks in apps today; and a surge of developers keen to get their hands on those APIs to build that functionality into their apps.

What Stream is working on is not to be confused with the customer-service focused services that companies like Zendesk or Intercom are building when they talk about chat for apps. Those can be specialized features in themselves that link in with CRM systems and customer services teams and other products for marketing analytics and so on. Instead, Stream’s focus are services for consumers to talk to other consumers.

What is a trend worth watching is whether easy-to-integrate services like Stream’s might signal the proliferation of more social apps over time.

There is already at least one key customer — which I am now allowed to name — that is a steadily growing, still young social app, which has built the core of its service on Stream’s API.

With just a handful of companies — led by Facebook, but also including ByteDance/TikTok, Tencent, Twitter, Snap, Google (via YouTube) and some others depending on the region — holding an outsized grip on social interactions, easier, platform-agnostic access to core communications tools like chat could potentially help more of these, with different takes on “social” business models, find their way into the world.

Stream’s technology addresses a common problem in product development by offering an easy-to-integrate and scalable messaging solution,” said Dick Costolo of 01 Advisors, and the former Twitter CEO, in a statement. “Beyond that, their team and clear vision set them apart, and we ardently back their mission.”

Continue Reading

Social

TikTok launches ‘TikTok Q&A,’ a new feature for creators to engage with viewers’ questions – TechCrunch

Published

on

Earlier this year, TikTok was spotted testing a new Q&A feature that would allow creators to more directly respond to their audience’s questions using either text or video. Today, the company has announced the feature is now available to all users globally. With the release of TikTok Q&A, as the feature is officially called, creators will be able to designate their comments as Q&A questions, respond to questions with either text comments or video replies, and add a Q&A profile link to their bios, among other things. The feature also works with live videos.

TikTok Q&A grew out of a way that creators were already using the video platform to interact with viewers. Often, after posting a video, viewers would have follow-up questions about the content. Creators would then either respond to those questions in the comments section or, if the response was more involved, they might post a second video instead.

The Q&A feature essentially formalizes this process by making it easier for creators — particularly those with a lot of fans — to identify and answer the most interesting questions.

Image Credits: TikTok

To use Q&A, viewers will first designate their comment as a Q&A question using a new commenting option. To do so, they’ll tap the Q&A icon to the right side of the text entry field in comments. This will also label their comment with the icon and text that says “Asked by” followed by the username of the person asking the question. This makes it easier for creators to see when scanning through a long list of comments on their video.

The feature will also feed the question into the creator’s new Q&A page where all questions and answers are aggregated. Users can browse this page to see all the earlier questions and answers that have already been posted or add a new question of their own.

Creators will respond to a Q&A question with either text or video replies, just as they did before — so there isn’t much new to learn here, in terms of process.

They can also add Q&A comments as stickers in their responses where the new video will link back to the original, where the question was first asked, similar to how they’re using comment stickers today.

The feature will also be available in TikTok LIVE, making it easier for creators to see the incoming questions in the stream’s chat from a separate panel.

Image Credits: TikTok

As a part of this launch, a Q&A profile link can be added to creators’ Profile bios, which directs users to the Q&A page where everything is organized.

During tests, the feature was only made available to creators with public accounts that had more than 10,000 followers and who opted in. Today, TikTok says its available to all users with Creator Accounts.

To enable the feature on your own profile, you’ll go to the privacy page under Settings, then select “Creator,” tap “Q&A” and then “Turn on Q&A.” (If users don’t already have a Creator account, they can enable it for themselves under settings.)

The feature is rolling out to users worldwide in the latest version of the TikTok app now, the company says.

@tiktokYou can now ask and answer any questions on LIVE with the new Q&A feature. Check it out now!

♬ original sound – TikTok

Continue Reading

Social

WhatsApp rolls out voice and video calls to desktop app – TechCrunch

Published

on

WhatsApp is rolling out support for voice and video calling to its desktop app, the Facebook-owned messaging service said Thursday, providing relief to countless people sitting in front of computers who have had to reach for their phone every time their WhatsApp rang.

For now, WhatsApp said its nearly five-year-old desktop app for Mac and Windows will only support one-to-one calls for now, but that it will be expanding this feature to include group voice and video calls “in the future.”

Video calls work “seamlessly” for both portrait and landscape orientation, and the desktop client is “set to be always on top so you never lose your video chats in a browser tab or stack of open windows,” it said.

Speaking of which, support for voice and video calls is not being extended to WhatsApp Web, the browser version of the service, at the moment, a spokesperson told TechCrunch. (Facebook launched dedicated desktop app for its Messenger service last year, which supports group video calls.)

The new feature support should come in handy to millions of people who use WhatsApp’s desktop client everyday and have had to use Zoom or Google Meet for one-to-one video calls on desktop partly because of convenience.

WhatsApp, used by over 2 billion people, hasn’t shared how popular video and voice calls are on its platform, but said it processed over 1.4 billion calls on New Year’s Eve — the day usage tends to peak on the Facebook-owned platform.

Like the 100 billion messages WhatsApp processes on its platform each day, voice and video calls are also end-to-end encrypted, it said.

Once known for taking quarters to push a feature improvement to its app, WhatsApp has visibly grown more aggressive with adding new features in the past year. In late January, Facebook added opt-in biometric fingerprint, face, or iris scan authentication for WhatsApp on desktop and the web, an additional protection layer that makes more sense after today’s update.

It rolled out ephemeral messages, photos, and videos that disappear after seven days late last year, and also rolled out its payments service in India, its biggest market by users.

The new feature additions come as WhatsApp is attempting to convince users to agree to its planned changes to privacy policy — which has received some heat on Tech Twitter. Whether those concerns raised by a handful of people on Twitter extend to the larger population remain to be seen.

Continue Reading

Trending