Investing in Realvision

Today, we’re excited to announce a new investment in Realvision. We see great opportunity in the crossover between AI and VR. The company’s machine learning and computer vision platform creates immersive 3D, VR-ready models for real estate listings. Not only does the platform create immersive models, but also creates an auto-generated floor plan and offers still photos to provide a comprehensive marketing package to enhance real estate listings, creating a better experience for prospective home buyers. Making the real world consumable in VR is a complex but valuable task. We look forward to working with Realvision to change the way we buy real estate.

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.

Feedback Loup: College Panel

We recently hosted a panel of 8 college students from the University of Minnesota. The goal was to better understand how millennials think about social media, communications, video, VR, AR, the selfie generation, the future of work, and privacy. Here’s a summary of what we learned:

Text Is Dying

  • Quote: “Texting replaced email, and photos have replaced text messages”.
  • Message: Text is being used less frequently by each of our panelists. They view text as a formal way to communicate. Snap, Facebook and Instagram are the preferred communication platforms, with Facebook settings being switched to photos only. The panelists mentioned tech platforms promoting messaging within games as a way to maintain usage.
  • Takeaway: Text is slowly going away, replaced by video and photos. Text is viewed more as a formal way to communicate.

Fake News

  • Quote: “I like Snap for news.”
  • Message: Our panelists get their news from a wide variety of sources. 7 of 8 panelists are not concerned about fake news. Snap was the most popular way to aggregate news from traditional sources (3 of 8), followed by mainstream news outlets; e.g., CNN and WSJ.
  • Takeaway: Professional news is still respected but not paid for by these college students.

The Future of Work

  • Quote: “It’s scary. If we can’t have cashiers, truckers and fast food jobs. . . how will people live?”
  • Message: College students know they are entering a workforce that will have dramatic changes over the next 30 years. They have concerns about who’s going to control everything as resources become more concentrated. The University of Minnesota offers a class titled “Size of the Future” that addresses the risk of job loss to automation. The group did consider these changes when thinking about a career, with an increased interest in a more technical education that feels more defensible. Ultimately these students believe that the negative impact of lost jobs will be partially offset by the positive impact of new industries being formed.
  • Takeaway: College students understand that the workforce is changing. They envision social challenges emerging from displacement of workers with lower levels of education. But they believe a college education will ensure that their futures are safe.

Read More

Apple Working With Tesla Is A Fairy Tale

There’s been a lot of talk about Apple buying Tesla, but what if Apple simply made a $10 billion equity investment in the company instead? It sounds so good — Apple working with Tesla. In theory, it would make our lives so much better. Imagine all of the things you love about your iPhone, perfectly integrated with all the things Tesla owners rave about. The two tech giants could take over the auto industry over the next 20 years as consumers embrace electric vehicles and automation. Unfortunately, an investment from Apple, nonetheless an acquisition, would be hard to pull off. At the end of the day, that might be better for consumers if not investors.

Before we discuss why it won’t happen, let’s go over why it sounds so good.

For Tesla. A $10 billion cash infusion would all but eliminate any current or future cash problems for the company. While $10 billion equity investment would cause about 20% dilution today, it’s likely it would have a long-term benefit on Tesla stock given the removal of the cash question. Aside from the cash, we believe Apple could and would want to provide resources from their world class hardware, software, and AI teams to make Tesla’s the entertainment system and autopilot better. The investment would likely remove Apple as a potential direct or indirect competitor. Additionally, Tesla’s Model 3 could be showcased in Apple’s 490 retail stores in 20 countries.

For Apple. Investors would feel like they are actually doing something with their cash, which should be a positive for AAPL’s multiple. Apple would be investing in a company that has the potential to be multiple times bigger over the next decade. They would not be spending on the impossible, which would be building its own car to try to catch Tesla, but rather investing in making the leader even better. The impact of AI and robotics on the automotive sector is one of the next mega tech trends, and Apple would have a pole position within that theme.

Read More

Tesla’s Bedrock In AI & Robotics Will Transform The Industry & Our Lives

I always wanted to cover Tesla, but as an internet analyst, the stock fell outside of my coverage space. Despite this, I continued to study the company and ultimately invested because I believe that Tesla is not a car company, but a consumer electronics company that thinks like an internet company. With a bedrock in AI and robotics, Tesla is one of the best positioned companies to transform our lives over the next 20 years.  We think Tesla is on par with Amazon when it comes to a reckless pursuit to shape the future, which we believe will reward investors over the long run.

The Street Is Understandably Focused On The Wrong Metric

Tesla reports December quarter results on Wednesday (Feb. 22). Given the 48% rise in TSLA shares over the past 3 months, now trading near an all-time high, it’s understandable why investors are nervous going into the print. After all, good news is priced in as information of the earlier-than-expected Fremont production retooling has stoked Model 3 production expectations. As of our last check, buy side investors expect 17k to 25k Model 3 shipments in 2017.  That’s a big number when you consider that in 2016 Tesla delivered 76k vehicles (all models) to customers. Investors will be zeroed in on Elon Musk’s comments on the earnings call about production of the Model 3 in 2017.  His comments may cause volatility in the stock short term, but they are irrelevant in the long run.

It’s Not About How Many Model 3’s Tesla Sell In 2017

As venture capitalists, we have the luxury of thinking about themes over a very long horizon. With that perspective, Wednesday’s Tesla earnings report is a non-event.  What’s more important is that Tesla makes the best car in the world, amplified by AI and robotics. That focus will keep competitors in check, allowing the company to reach scale and ride the next tech mega wave as our lives are quickly transformed (over the next 20 years) into an electric, automated existence.

Artificial Intelligence

Tesla’s obvious AI play is autopilot for autonomous vehicles, with a less well known AI push in manufacturing. We know that the company is pushing boundaries to gain data to improve its driving AI with a goal of being first to market with an L4 compatible vehicle (the automated system can control the vehicle in all but a few environments).

The first to market will have a measurable advantage because road data equates to smarter AI and safer cars.  Google’s Waymo has driven over 2 million autonomous miles, but comparisons with other automotive companies are difficult given some companies include simulation miles.  Last October, Elon Musk reported Tesla had driven 222 million cumulative autopilot miles, but those miles are not comparable to the fully autonomous number that Waymo reports.  It’s unlikely that Waymo will have a commercially available vehicle in 2019, but likely that Tesla models solid in 2019 will be L4 compatible. Traditional automotive is even further behind, with BMW, Audi, Mercedes, Ford and GM likely shipping L3 autos in 2019. Note that L5 is the highest level of autonomy, for vehicles capable of all aspects of the dynamic driving under all roadway and environmental conditions that can be managed by a human driver, followed by L4, L3 and so on.  This begs the question, why would anyone interested in an autonomous car buy an L3 compatible vehicle if it was priced similar to an L4 vehicle? We don’t know how Tesla’s autopilot AI stacks up against the market, but based on comments from our industry contacts, Tesla sees AI as one of its two core competencies and is structuring its future around it.

Read More

Face Off: Amazon Echo vs. Google Home

As digital assistants continue to improve, more and more users are integrating them into their daily routines. In our Robot Fear Index, we found that 43% of Americans had used a digital assistant in the last three months. But we believe that we are still in the early innings of how natural language processing will improve our daily lives and our interactions with machines.

To see how far along our digital companions are, we chose to review the two most popular home assistants: Amazon Echo and Google Home.  We put the devices to the test by asking them 800 different everyday queries.  Google Home came out on top, answering 39.1% of the queries correctly vs. the Echo at 34.4%.

Methodology

For this experiment, we asked the same 800 queries of both the Echo and Home.  We graded the queries on two metrics:  First, did the device understand what we asked correctly?  Second, did the device answer the query correctly? In our study, Amazon Echo understood 94.4% of the queries we asked and answered 34.4% of all queries correctly. Google Home understood only 77.0% of the queries we asked, but was able to answer 39.1% correctly.

In our study, Amazon Echo understood 94.4% of the queries we asked and answered 34.4% correctly.  Google Home understood only 77.0% of the queries we asked, but was able to answer 39.1% correctly.

One reason that the Amazon Echo had a higher rate of understanding queries was due to our ability to confirm the data using the companion app.  This app gives the user a live feed of what Amazon Echo heard.  Google Home does not offer a transcript of what the device picked up.  Because of this, it was difficult to tell if Google Home understood the queries but couldn’t answer them, or if it truly had a harder time understanding queries. Since we were unable to see exactly how well Google Home understood our queries, we assumed that if Google Home responded that it was unable to perform a certain function, then it had understood the query correctly.  For example, if we asked “Hey Google, send a text to John” and received a response “Sorry, I can’t send texts yet,” then the query would be marked as understood correctly, but answered incorrectly.

Results

The queries that we asked divide into five buckets: local, commerce, navigation, information, and command.  When breaking down the comparison of correct answers between the Echo and Home, Amazon predictably performed better at commerce and Google at information, in-line with their stated missions.

Read More