Amazon has added basic voice recognition and personalization features to its Echo devices, as a partial response to the Google Home’s similar feature. As in other areas, Amazon has a weakness here compared with Google in that it has no real background profile information on the individual users in a household, something it’s starting to change with recent family features (and the teen account feature announced earlier today). As such, its voice recognition feature will only enable limited personalization, focused on Amazon’s own services and not third party features like calendars, which is where Google Home’s equivalent feature (and Google services in general) excels. This makes Amazon’s new feature a good start, but far from a fully-fledged response to Google in this department, while it continues to be ahead in other key areas following its recent hardware and software upgrades.
via The Verge
Chinese e-commerce giant Alibaba has announced that it will double its R&D spending over the next few years, with the main focus of its $15 billion investment in three areas: AI, the Internet of Things, and quantum computing. Though the Bloomberg piece here suggests this is another front in its competition with Amazon, the reality is that the two compete for very little business, mostly competing against other western and Chinese companies respectively in their major markets, with Alibaba dominant in China but negligible in the US. Rather, this is part of a much broader competition between big tech companies in aiming for leadership in several key new research categories, AI chief among them. That’s a competitive dynamic that has now taken on geopolitical overtones, with the US and China emerging as the likeliest leaders in AI, itself considered critical to future competitiveness in not just business but cyber warfare and other areas. Alibaba is, of course, already growing dramatically as a company, and the increase in planned R&D spending is therefore not wildly out of whack with its overall growth, but it makes for a great press release as the company seeks to burnish its innovation credentials.
Microsoft is bringing its Cortana virtual assistant to Skype, over a year after it first demonstrated some of the features at its Build developer conference in 2016. Whereas Cortana does act in some settings as a voice assistant like Siri and the Google Assistant, it’s worth remembering that Microsoft uses the Cortana name to refer to all the underlying AI capabilities too, and that’s what’s being implemented here. The integration is text- rather than voice-based and limited to messaging rather than voice or video interactions, and Cortana will offer up smart replies in messaging conversations and also offer useful information like movie and restaurant reviews. In some ways, this is a different spin on the Context Cards Snapchat added today, and very much along the same lines as Facebook’s current implementation of its M assistant within Messenger – offering context-based suggestions within existing human-to-human interactions. This is part of Microsoft’s broader push to get its AI into every corner of its products and services, but will hampered by the overall stagnation of Skype as a communication platform – though it clearly has some messaging users, it isn’t the default messaging platform for the vast majority of the population, at least in their personal lives.
I haven’t seen an official announcement around this, but Windows Central reports that Microsoft has quietly added support for four smart home vendors – Nest, SmartThings, Hue, Wink, and Insteon – to its Cortana virtual assistant. On the one hand, this is good timing with the Harmon Kardon speaker apparently getting ready for launch, but on the other it’s odd given the recent voice assistant partnership between Microsoft and Amazon, a big selling point of which was being able to control smart home gear through Alexa. In fairness, the latter still has much broader support for smart home ecosystems than Cortana, but Microsoft’s assistant now talks to several of the largest, and these plans must have been in the works for months now, certainly before the Alexa partnership was announced. At any rate, it’s going to be much simpler to control these devices directly through Cortana than through the awkward two-step process the Alexa partnership would require, and this is a good addition ahead of the launch of Cortana-based speakers.
via Windows Central
Quartz reports that Alphabet’s DeepMind subsidiary, which is still registered as a separate private company in the UK and therefore has to report its own financials, lost $162 million in 2016, on revenues of just $40 million, all of which came from Google. It’s a quirk of accounting that DeepMind is still reporting as a separate company, but it gives some insight into the cost of running such a business, which is focused on cutting-edge AI work, much of which is not ready for direct monetization in revenue generating products. Given that Alphabet as a whole spent over $15 billion on Research and Development in the past year, this is a tiny fraction of the total, and an operation the company can easily afford to keep going along these lines. Much of the losses, incidentally, stem from the $137 million the company spent on staff and related costs, of which I would guess a big chunk is stock-based compensation, which runs at $2 billion per quarter for Alphabet as a whole and $100-150 million per quarter in the Other Bets segment. And of course there are big chunks of Google itself working on AI as it relates to specific products too, so this is far from the scale of Alphabet’s overall investment in AI, which is increasingly filtered into everything Google does.
Here’s a roundup of some of the smaller announcements Google made today, including the Pixelbook Chromebook, PixelBuds wireless earbuds, and an intriguing AI-powered camera called Google Clips. The Pixelbook is true to the original Pixel Chromebook from Google, which was equally bizarrely positioned as a premium device in a category which is mostly appealing for its low cost. It’s added some hot recent trends like convertibility and a $99 pen, ChromeOS has added Android app support, and Google is debuting its Google Assistant on a laptop here as well. None of that is likely to overcome the inherent funky positioning of a $999-plus Chromebook, and it’ll continue to be a marginal device. That Google should continue to compete here rather than entering the smartwatch market directly feels funny given how much more the Android Wear ecosystem needs first party hardware from Google than ChromeOS does.
The PixelBuds earbuds are in the “neckbud” category rather than the truly cordless earbud category Apple’s AirPods dominate today, and I think that’s fine – I’m wearing BeatsX on a plane as I write this, and continue to like these better than AirPods, and I think this category has a lot of value. The earbuds are priced the same as AirPods, and as with those buds, come with a voice assistant built in, though Google’s big differentiator is real-time language translation, which was successfully demoed on stage. Of course, most of us only rarely (if ever) need such a function, so this is more of a gimmick than a useful feature for now, but it’s a great gimmick.
Lastly, Google’s big surprise at today’s event was one of the last things it unveiled, which is a small standalone camera which is designed to unobtrusively capture pictures and video in the home, powered by AI which will determine when and how to take them. That’s a brand new concept, though it obviously competes to some extent with both Samsung’s Gear 360 line and cameras from the likes of GoPro, whose stock took a big hit today. In reality, of course, this product likely won’t sell in any big numbers because the category doesn’t exist, because it’s priced at $250, and because Google doesn’t have the presence or history in hardware to launch a new category, and it’s best seen – like the real-time translation feature in the PixelBuds – as evidence of Google’s AI chops, and as something which might therefore come to other Google products in time and thereby reach a broader audience.
Google today announced both larger and smaller versions of its Google Home device, while adding software features to its existing hardware, as part of its second generation hardware launch event in San Francisco (see here for my comment on the Pixel 2 smartphones it also announced). It’s a busy time for voice speaker announcements, coming a week after Amazon’s big update of its Echo line, and the same day as Sonos’s voice speaker launch, but we now have a much clearer picture of how the lineups of major vendors will be positioned to finish out the year and going into next year.
Amazon has a pretty mature product line now, but still no direct entry in the premium audio space, a segment it seems willing for now to cede to partners and competitors. Apple is entirely focused on the high end market, with its HomePod priced at $349 and coming in December, while Sonos is trying to find a niche between these two markets with its $200 Sonos One speaker and a neutral approach to voice assistant and music ecosystems. Lastly, we now have Google pursuing a good, better, best strategy like Amazon, but with its best much more focused on premium audio than Amazon’s new Echo Plus, which seems more geared towards smart home support and costs far less.
It’s fascinating to see Google come in above Apple in its pricing for the Google Home Max, at $400, suggesting it’s not going to be dragged down the pricing slide with Amazon but wants to make real margin on its products in the category. Given how much complaining I’ve seen about Apple pricing itself out of the voice speaker market, this new announcement certainly adds an interesting wrinkle. Of course, Google is also providing a cheaper speaker at $50 to compete more directly with the Echo Dot from Amazon, and is smartly focusing there as in its core Google Home product on design which will fit much better (and more subtly) in a home environment. Google should take significantly more share than it did last year with this new range of devices, especially the Mini, and it already took decent share with the first generation products. All in all, this is a great set of announcements from Google that should do pretty well, with the possibility of more to come in the speaker-with-screen segment early next year.
Amazon has acquired Body Labs, a startup which makes software for creating detailed and realistic 3D maps of people’s bodies, for somewhere between $50 and $100 million. The technology has an obvious connection to Amazon’s Echo Look, one of its more marginal Echo devices, but one which has potential to drive strong ties with the burgeoning clothing side of Amazon’s e-commerce business. Beefing up the capabilities of that device and the associated app-based capabilities for evaluating fashion looks and the like could therefore pay off in a big way for Amazon as it looks to differentiate itself from other clothing retailers.
Apple Has Acquired a Small French Photo Analysis Company (Sep 29, 2017)
Apple has made another one of its characteristic quiet, small acquisitions of a technology company, this one a French business specializing in computer vision for photo analysis. Unlike some other photo analysis tools, however, this one isn’t so much about recognizing the content of photos as determining which photo in a group might be technically best, or which photos are duplicates. It’s easy to see those technologies being used in future version of Apple’s Photos app on the iPhone to select the best picture from a burst of photos, or to manage a photo library on a Mac. Apple has put enormous attention into its cameras almost from day one of the iPhone, but its photo management software hasn’t kept pace for much of that time, though recently it’s began to invest more seriously in it both on the iPhone itself and on the Mac. This small acquisition is a sign that it plans to continue to make incremental improvements, if nothing else.
Amazon today held what many publications described as a “surprise” event (in that Amazon embargoed the very existence of the event) to announce broad updates to its Echo line of devices, as well as a new version of its Fire TV box. The announcements represent a maturing of the Echo product line, which went from three main entries to five, now with a good, better, best approach to pure speakers and small and large options for speakers with screens. I’ve just created this image for the column I’m writing for Techpinions for tomorrow, and it’s a good overview of the Echo lineup before and after today’s announcements. Amazon also announced two new accessories: the Echo Connect, which acts as a bridge between an existing landline phone and Alexa calling, and Echo Buttons, the first of a new category of accessories called Alexa Gadgets, which will serve as companions to Echo and other Alexa-enabled devices, offering additional functionality (the Buttons are envisaged as interfaces for family members playing voice games, for example).
What we’re seeing from Amazon here is a consolidation of its early leadership in the voice speaker category, re-emphasizing its desire to dominate that market, if necessary through pricing hardware at or below cost. It engaged in some clever positioning around the pure speaker space by moving its core Echo product down in price by $50 while significantly improving its industrial design and audio performance, and introducing a new tier at $150 under the Echo Plus name. The Echo Plus also serves as a smart home hub in its own right rather than merely using cloud services and APIs to control devices through existing hubs, which is an interesting step forward but will require smart home gear to integrate with it in new ways. Amazon also announced Alexa integration in BMW cars from the 2018 model year onwards and Minis from mid-2018 onwards, which is another step in taking Alexa out of the home, albeit one which will take many years to reach a meaningful proportion of cars on the roads. Lastly, Amazon updated its Fire TV box, now in a quasi-dongle form factor, with 4K and HDR support and an Alexa remote (but not the always-on feature in the box itself which had been rumored), and at a slightly lower $70 price.
Both the timing and content of Amazon’s announcements today are a big thumb of the nose towards Google, which of course is holding its fall hardware announcement next week, and in the context of the secrecy around today’s event, I wonder if Google got wind of it yesterday and decided to rain on Amazon’s parade ever so slightly with its yanking of YouTube from the Echo Show. The only big move in the voice speaker space we’re expecting next week from Google is a smaller device to compete with the Echo Dot, so Amazon just wiped the floor with those announcements, making its own hardware more price competitive even at list price and adding new options for discerning customers. All of this also makes life a little tougher for both Sonos and HomePod, with Sonos announcing its first voice-enabled hardware on the same day as Google’s event next week. Audio performance on basic voice speakers is now getting good enough that both Apple and Sonos need to demonstrate significantly superior performance and better experiences with multi-room audio to compete.
The Financial Times reports that Google is working on an AI-based tool that will help publishers identify possible subscribers for their newspapers. This is a somewhat fleshed-out version of a report from a month ago on Bloomberg, which had fewer details but said Google was testing a number of different approaches. As a reminder, the context here is the tension between news organizations and both Google and Facebook over business models, the increasing power of the internet companies, and the challenges of selling online subscriptions and building brands when search and social serve as channels for so much news consumption. As I’ve said before, Facebook began taking this tension seriously some time ago and pouring oil on troubled waters, but it seems to have taken Google longer to come around, and it’s still mostly at the testing stage in its efforts. Putting AI to work in the service of solving the problem is a classic Google move, but it remains to be seen how effective that will actually be. Certainly, the publishers quoted by the FT seem heartened but not yet won over by Google’s new approach.
via Financial Times
Facebook Opens AI Lab in Montreal (Sep 15, 2017)
This is a minor thing, but nevertheless an important one in several ways. Apple has updated the executive bios on its website to reflect a few changes, notably the change in responsibility for Siri from Eddy Cue (generally responsible for online services) to Craig Federighi (responsible for software), and Eddy Cue’s ownership of Apple’s original video content push. That’s notable for two reasons: one is that Eddy Cue has lost other areas of responsibility recently, notably the App Store to Phil Schiller, and Siri is an area where Apple can ill afford to be seen to be falling behind the competition. Taking it away from Cue is likely a sign that Apple wants to see the same rapid improvements there as it did in the App Store when Schiller took over, but also a recognition that the content push is going to take more of Cue’s attention going forward.
Also worth noting: though there’s still only one woman among Apple’s top-tier leadership of SVPs and CXOs as shown on its executive leadership page, the next tier of VPs is now half women, with three of the four women of color. Diversity in the top ranks at Apple has been poor and slow to change, in part because the senior leadership team has been so stable for so long, but it’s clear that Tim Cook is using the more frequent changes happening at the next tier down to increase diversity there.
via Mac Rumors
★ Amazon and Microsoft Announce Cortana-Alexa Integration (Aug 30, 2017)
Microsoft and Amazon have officially announced that their respective assistants will begin working together later this year, news broken by the New York Times along with interviews with the companies’ CEOs. Of the four major voice assistants, these two are arguably the weakest, for all that the prevailing narrative is that Amazon is ahead in voice. As a reminder, Amazon has perhaps 15-20 million users of its Alexa assistant today, while Microsoft has 145 million regular Cortana users, Google has hundreds of millions of Android devices in the market with some form of its voice assistant technology, and Apple has nearly a billion Siri-enabled devices in use, with 375 million monthly active users as of June. More importantly, both Amazon and Microsoft are bound to a single category of devices today: home speakers for Amazon and PCs for Microsoft. Yes, both have smartphone apps too, but they’re very much second class citizens behind the built-in assistants available from the lock screen on the two major smartphone platforms. So the coming together here makes a certain amount of sense on that basis.
However, this doesn’t solve that fundamental problem of getting first party status on smartphones, and the integration the companies will offer will at least at first be both awkward and limited. Users of either assistant will have to invoke the other using double commands (“Cortana, open Alexa…” or vice versa) before even speaking their request. The integration itself will likely focus on smart home control from Cortana and personal information management through Microsoft’s apps from Alexa, filling an important gap in Amazon’s portfolio given that it lacks its own broadly-used calendar, contacts, reminders, or other PIM apps. In theory, the integration will get less awkward at some point down the line, with each assistant deciding on the fly which underlying AI to use to process a request, but in practice that seems challenging.
For today, it’s relatively straightforward given that the two assistants excel in different domains, but Microsoft’s partners are about to launch the first Cortana-powered speakers and other home devices that will compete more directly with the Amazon Echo, and the overlaps between their capabilities will only grow over time. So who will decide which AI handles which requests? Will this integration only live as long as the companies can agree on that? Or will the lead assistant in each case grab the tasks it wants and leave the dregs for the other? Meanwhile, both Google and Apple will make inroads into the home speaker space in the coming months, allowing them to provide more ubiquitous voice assistants and erode Amazon’s early lead in the home voice market. To summarize, though the logic behind a deal here is reasonably sound, it’s likely to be strained over time and less relevant as the two larger voice platforms expand in the home.
Note: for non-subscribers, I’ve temporarily opened access to the “Amazon is Ahead in Voice” narrative evaluation linked below, so you can go and read (or watch a video on) the broader context for this move and why I say above that Amazon is one of the weaker rather than stronger players in this market.
via New York Times