What Google I/O Means for Immersive Computing

This week Google hosted it’s annual I/O developer conference. On day one, the company focused on their innovation in the artificial intelligence space. On day two, they talked about new VR products. Here’s our take on what the latest out of Mountain View means for the future of AI and VR.

  • Day One: AI. At Google I/O, Google lived up to it’s commitment to be an AI-first company. The company announced a slew of AI innovations focused on making their platform easier to use with more natural interfaces including voice (Google Assistant) and vision (Google Lens). For example, Google Assistant now includes support for calendars, phone communication, and proactive alerts, closing a gap we identified in our work on home assistants. Proactive alerts for voice-based assistants is a big step towards a screenless future. Google Home now flashes when it has relevant and timely information. For example: [*Google Home flashes*] “Traffic is heavier than usual. Leave in the next five minutes to be sure you make it to Anna’s soccer game on time.” In the screenless future, friction-less information push represents the future of search technology. Google is still in the best position to own the category given its organization of the world’s information. Google’s progress in the fields of computer vision (Google Lens) and cloud-based supercomputing/machine learning (Google Compute Engine) positions the company for success as we transition to more natural and immersive computing. It’s no coincidence that day one ended with a tease for a standalone VR headset, untethered to a PC or a smartphone. For more, see the 10 min condensed version of all the day one announcements here.
  • Day Two: VR. The big news on day two was the announcement of a standalone VR headset untethered to a PC or smartphone for computing power. Google is partnering with Qualcomm to build a reference headset and announced partnerships with HTC and Lenovo to bring standalone VR headsets to market later this year. Google also addressed a common knock on VR: given the full enclosure of VR headsets, VR experiences are hard to share with others. Google is making VR more social with shared rooms and voice chat as replacements for the text-based comments familiar to PC users. These advancements will help make VR mainstream faster. The transition from PC- and smartphone-driven VR to standalone VR will take 3-5 years (we don’t expect real traction – 1m units – until 2019), but the transition has clearly begun.

Bottom line: Google’s investments in AI and VR will accelerate the transition from computing on PCs and touchscreen devices into the future marked by immersive computing.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Apple Acquires AI Company Lattice Data

Apple recently acquired Lattice Data, a dark data mining company. Lattice Data takes “dark data,” meaning unstructured text and images, and turns it into structured information that can be used by traditional data analytics tools. Lattice Data began as a Stanford research project called DeepDive, which was also an attempt to process dark data on the internet.

An estimated 70-80 percent of data on the internet is unstructured, dark data. Artificial Intelligence can be used to structure this data, making it more useful. While Apple has not made it clear how it will use Lattice Data’s technology, we expect the company to leverage Lattice technology to improve Siri. For example, by accessing this dark data that is currently unusable, Siri could become much more effective in our daily lives. Siri would be able to answer our obscure questions about odd historic events, have a better understanding of which movie theater we are talking about, or even find the exact hot dish recipe that we are looking for. Eventually, much of the user-generated content on the web could be accessible to digital assistants.

Lattice Data falls right in line with Apple’s acquisition playbook: Buy companies that have a kernel of something special that Apple can put its significant resources behind to build into something even more amazing. They did it with P.A. Semi to build better processors for the iPhone. They did it with Siri to build Siri. They did it with Authentec to build Touch ID. They used Beats, both the executive team and the technology, to build Apple Music. In these instances, Apple took new technology and incorporated it into its platform soon after. It was clear that Apple was making an investment in a core technology. And with Lattice Data it’s clear that Apple is making an investment in AI.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

An Update on Commercial Drones #AUVSI2017

If artificial intelligence is the brains, then robots are the brawn in the future of work.

Earlier this week, we attended the AUVSI XPOENTIAL trade show in Dallas, the largest global gathering of unmanned systems providers, robotic software developers, and industry experts.

We sat down with 13 executives from some of the leading commercial drone companies. Here are the takeaways from our meetings:

  • We’re Still in the Early Innings of Drones. Overall demand for drone hardware, software, and services continues to see strong momentum domestically and internationally. Industry leaders are confident that the market potential in core verticals remains deeply unsaturated. Increasing customer awareness around the efficiencies of drone technology represents a catalyst for accelerating industry growth.
  • Ground Robots and Drones are Complementary. Terrestrial robots and aerial robots (i.e., drones) should not be considered competitive offerings. Some of our favorite strategies in the space couple ground robots with drones for autonomous services.
  • Drone Delivery is Real. Throughout our meetings, robotic delivery of packages was identified as one of the largest untapped markets. Most industry leaders do not expect regular drone delivery to be viable for 5+ years; however, that’s a meaningful uptick from last year, when a similar group did not expect drone delivery to be realistic for 10+ years. Amazon is leading the way in the drone delivery space, but will likely acquire enabling technologies to make it a reality.

Read More

Tesla is the Next Apple

This note was originally published on Business Insider.

Apple is the world’s largest company with a market cap of nearly $770 billion as of this writing. Tesla is one of the world’s largest automakers with a market cap of close to $55 billion, although we think the Tesla story is just getting started.

There are many parallels between Apple about a decade ago and Tesla today, market cap being one of them.  In Q4 2005, Apple’s market cap was close to where Tesla’s is today ($54 billion). A decade from now, we think we’ll look back at Tesla and realize it was the next Apple.

There are five major similarities to Tesla today and Apple in the mid-2000s:

  1. Brand
  2. Visionary leadership
  3. Integrated hardware and software
  4. Halo effect
  5. Reshaping a market

Brand

Tesla’s brand is to the car industry what Apple’s brand is to consumer electronics. Tesla owners love their Teslas.

According to a Consumer Reports survey, 91% of Tesla owners state they would “definitely” buy their cars again, the highest rating of any automaker. The next two closest automakers were Porsche at 84% and Audi at 77%. By comparison, Tim Cook stated on Apple’s Dec-16 earnings call that iPhone had a 97% satisfaction rate.

Beyond tangible customer satisfaction metrics, we believe there is a less tangible cool factor to the Tesla brand, much like Apple in the late 1990s and early 2000s. In many ways, Tesla has the same “think different” attitude that Apple popularized.

Tesla has built a brand around being a different kind of automaker. Not only because its vehicles are powered entirely by electric, but also because they don’t use model year numbers and treat software updates more similar to updating an iPhone app than a car. The company has done this all while squarely placing itself in the conversation with BMW as one of the best-engineered cars in the world. Tesla has established itself as an aspirational brand by taking a new approach to the car market.

Visionary Leader

Elon Musk and Steve Jobs share similarities in that they are visionary entrepreneurs that simultaneously operated multiple groundbreaking companies.  Musk with Tesla and SpaceX and Jobs with Apple and Pixar.  However, both seem to have different guiding lights.

Read More

Face Off: Siri vs. Google Assistant vs. Cortana

The importance of voice assistants in the screenless future is hard to overestimate. We see speech-driven user interfaces as a key component of the next computing paradigm, so it’s helpful to understand empirically where each platform stands today. In February, we compared Amazon Echo and Google Home to see which assistant was winning the race to become the centerpiece of the home. Google Home won that battle, but it was close. For our next digital assistant face off, we tested the three most prevalent digital assistants available for mobile devices: Siri, Google Assistant, and Cortana. This time, Google Assistant came out on top.

Methodology

We asked the same 800 queries to each assistant that we asked Amazon Echo and Google Home. We graded the queries on two metrics: First, did the assistant correctly understand the query? Second, did the assistant answer the query correctly?

The queries break down into five categories:

  • Local – Where is the nearest McDonald’s?
  • Commerce – Where can I buy more printer paper?
  • Navigation – How do I get to REI from here?
  • Information – What is Apple’s stock price?
  • Command – Remind me to call Mom at 2pm today.

Results

Google Assistant, the clear winner, understood 99.9% of the queries we asked and answered 74.8% of them correctly. Siri understood 94.4% of the queries we asked and answered 66.1% of them correctly. Finally, Cortana understood 97.3% of the queries we asked and answered 48.8% of them correctly.

By category, Google’s lead in navigation and information is demonstrable, but there’ more parity between Siri and Google Assistant in local, commerce and command related queries. Cortana lagged both Siri and Google Assistant in all categories, narrowly coming in second only in information.

Read More