SITALWeek #193
Stuff I thought about last week 5-19-19
Click here to SIGN UP for SITALWeek’s Sunday EMAIL
Greetings – the rising structural costs and culture clash of doing business with China; the pitfalls of the misguided “Long Term Stock Exchange”; gaming rivals Microsoft and Sony partner as the cloud becomes vital for every industry; the damage to Facebook from a potential FCC consent decree; and much more in this week’s newsletter - as always respond back with any thoughts or comments.
[Please read important updated disclaimers at the bottom of this post]
Stuff about Innovation and Technology
Google announced the Translatotron, which translates from one spoken language to another without intermediate text translation, making it faster and able to capture cadence/intonation of speech. Even more mind-boggling, Translatotron can even use a model of your voice so the resulting translation is in your actual voice! More info and audio samples of the new AI in action can be found here. That’s one more box checked off on Gene Roddenberry’s predictions of future tech from Star Trek - we’re still waiting on the transporter and faster-than-light travel – scientists need to get back to work!
Robot-picked apples are landing in grocery stores: “Abundant’s picker has more in common with a really smart Hoover vacuum than a human hand. The robot moves down rows of orchards and uses artificial intelligence with a dash of LIDAR to search for ripe apples. Once spotted, a robotic arm with a vacuum gently sucks the apples from the tree into a bin.”
San Francisco has banned the use of facial recognition technology by city agencies, although private companies are still free to use it in the city. And, in what appears to be evolving into a weekly feature of SITALWeek, this week’s questionable use of facial recognition involves the NYPD uploading a picture of Woody Harrelson based on a witnesses description.
The “digital twin” – a technique that allows complex simulation of real-world designs in software – is a topic I’ve covered in the past. Recently, Siemens announced a digital-twin system for chips designed for autonomous driving, which allows you to model the entire chip design and how the chip would perform in an autonomous vehicle in real-world simulations. This is just one type of solution enabling auto makers to start designing their own chips, which is part of the big trend of verticalization in semi design, much like we have seen with Google’s AI chips and Tesla's autonomous chip.
Whole Foods, Starbucks, and other retailers now allow tap and pay with cryptocurrencies. While this advance seems cool, it’s interesting to note that any innovation in US payments still has to ride the legacy and expensive rails of the outdated credit card and bank oligopolies. In contrast, China’s Internet gaming, social media, and payments powerhouse, Tencent, had the following to say on their earning’s call this week:
“...if you look at credit card charges in China, it's actually much lower than credit card charges around the world as well. I think it's really because of the fact that the Chinese economy was actually built -- the payment plan infrastructure was actually built at a later time and as a result, it's somewhat reflective of a lower cost. If you think about the credit card charges, then you will determine that it was actually a long time ago in which you pay a much higher IT cost, you may pay much higher communication costs. But today, the efficiency of the entire system is actually higher. So to some extent, I felt that the Chinese credit card charge is actually a good benchmark of what the cost it is today...”
Experimental hearing technology uses electrodes and deep-learning to select the voice you want to hear and block out other noise. While this particular technology may seem a little invasive, it’s much easier to imagine the feasibility of even a simple pair of glasses with a camera focused back on your eyes – such glasses could match your gaze with stereo mics and then process the sounds on your phone and send it focused back to wireless earbuds. (And, such technology could also mute out sounds, or people, that you don’t want to hear!)
Amazon will pay your salary for three months and cover $10,000 in startup costs if you quit working at Amazon and start a delivery company – a new stimulus for its existing delivery service partner program. In other Amazon shipping news, the company is building an extensive European air network like they did in the US.
A couple of weeks ago, we found out Gen Z loves malls, and now we discover that Amazon is buying and repurposing older defunct malls (video) for logistics hubs – who said malls were dying!?
Zillow CEO Rich Barton on the pitfalls of a founder returning to a CEO position: “Team members are less likely to challenge the founding CEO, so he can get himself and the company in trouble if his ears are not as big as his mouth.” You can read more on Zillow’s iBuyer business model in this WSJ interview, including the value of bringing liquidity to an illiquid market.
We’ve seen some crazy “co-opetition” in the tech world over the years, but mega video game platform rivals Microsoft and Sony teaming up to deliver gaming and AI experiences might be one of the biggest odd couples. And, why would this be happening? In short, the scale of the cloud and the network effects of data-driven AI are re-mapping the entire economy with a new digital operating system. As a result, Sony can’t afford to spend $10B+ on capital and R&D to build what they can get from Microsoft. The real question here is: why did Sony choose Microsoft when more neutral options exist? Google has announced an audacious gaming platform effort and Amazon (I suspect) will soon announce something similar. Another twist on this announcement is that it involves Sony’s advanced image sensors (a little known division that’s a big profit driver, making cameras for iPhones, etc.), with the implication being that Microsoft will work with Sony for its HoloLens enterprise augmented-reality goggles.
Disney this week took exclusive control of growing streaming giant Hulu, which will generate significant opportunities to bundle Hulu with Disney+ and ESPN+. Further, the deal generates technology scale, advertising scale, and data that will allow Disney to connect directly with consumers and offer merchandise, parks, travel, and other House-of-Mouse experiences. Disney can now also more easily program originals for the Hulu platform to help grow the subscriber base even faster. In related news, Disney’s Hotstar streaming service in India (the largest in the country) recently had a global record breaking 18.6M live concurrent streams of a Cricket match. Hotstar will also be launching other Disney streaming services to India. One last point on Hulu: the Hulu CEO did an informative interview on Peter Kafka's podcast last week discussing new ad formats and potential for the service. Twist my arm and I’ll throw in one stat on the power of the Disney content – National Geographic is the 14th biggest account on Instagram at 109 million followers!
Back when Google faced the existential threat of usage shifting from desktop to mobile, I always said “well if they don't make the transition to phones, they can always put an ad on their homepage”... From Google’s ad summit this week: “The search giant on Tuesday announced an expansion of its advertising real estate to boost revenue from mobile shoppers. It will feature ads on the homepage of its smartphone app worldwide, show more ads in Maps and place ads with image galleries in search results.”
News this week that the FCC is considering a 20-year, privacy-related consent decree against Facebook is likely a bigger deal than most people think. Consent decrees can be a prescription for corporate paralysis. Microsoft took more than ten years, a complete cultural shift, and a new CEO to get past their consent decree related to the bundling abuse of Internet Explorer and the Windows desktop operating system. Back in 2010 Google faced an investigation over the acquisition of mobile ad network Admob, (which was accidentally approved without a decree because the FTC said Apple was going to a mobile ad competitor – oops!), and then a consent decree over their acquisition of flight data company ITA (requiring Google to license ITA data to travel search competitors). While neither were onerous outcomes, Google became very cautious around M&A and verticalization product decisions for several years (lately, the shift to voice search and AI assistants has largely outweighed those fears at Google). If every little product decision at Facebook requires an oversight panel, I believe it will be very difficult for the company to carry out Zuckerberg’s vision of merging and re-casting Facebook, Messenger, Instagram, and WhatsApp. In February 2017, Mark Cuban made the argument at a conference I was attending that the US can’t break up the big Internet platforms because it would open a window for China to pull ahead globally...fast forward 27 months and Facebook COO Sheryl Sandberg makes an identical argument to keep Facebook together, which, coming from her at this moment, seems like attacking a strawman. Yet, with the potential for runaway network effects in AI, the US has a lot to lose to China if we materially handicap the US platforms.
Germany startup Lilium flight tested its five-passenger electric vertical-take-off plane, which is powered by 36 small motors and has a potential speed of 186 mph.
Samsung announces technology to get to 3 nm semi manufacturing in 2021 (and eventually 1 nm – for reference, a human hair is 2.5 nm), likely a year ahead of Taiwan Semi and several years ahead of #3 Intel.
China claims a major breakthrough in lithium processing that costs ~1/6th of the current wholesale price per ton. If true, it’s a huge boon to EVs, but would lock China in as the dominant lithium supplier.
Miscellaneous Stuff
Black Mirror Season 5 drops on June 5th and I already have anxiety just watching the trailer.
If you want to geek out for 16 minutes on quantum hacking, here is a good video from Minute Physics explaining how quantum computing can break all of our current encryption (if we ever find a way to get the information out of one without collapsing the information itself!). The key is that a quantum computer can do a superposition and calculate all possible answers simultaneously for a key input into the encryption cracking code.
Stuff about Geopolitics, Economics, and the Finance Industry
With technology deflation driving sustainably low interest rates for years to come, the whole world remains drenched in an abundance of capital. Were that not the case, the alarm bells would be crescendoing for a massive venture capital bubble. The numerous private companies that have taken dangerous amounts of dumb money as an existential subsititue for building real businesses are still mostly isolated from the rest of the economy. However, there will be some public tech stock impacts if a shock causes the non-traditional VCs like Saudi Arabia to back out of Silicon Valley. For sure, the online ad business at FB and Google will suffer as these dead-man-walking businesses stop buying ads, and we will see other knock-on effects, like an air pocket in cloud spending (and thus cloud CapEx, GPUs, etc.), but overall much bigger demand trends are driving the majority of public tech stocks. In the meantime the bubble remains strong as evidenced by the q&a site Quora raising funding at about 100x revenue.
Obligatory trade war comments: I think the reality is setting in with the markets and businesses that we are simply in a long-term period where the cost of doing business with China will stay elevated. News this week of a new government approval required to sell components to Huawei is crippling news for the company, which derives 1/3 of its components exclusively from US suppliers such as Intel, Xilinx, etc. This move also ropes in other countries, in particular Germany and the UK, which have approved Huawei 5G gear for telecom carriers to deploy. Ray Dalio smartly suggested this week that the trade war is a substitute for a broader cultural war: East vs. West, collectivism vs. individualism, equality vs. liberty, totalitarian vs. democratic, and so on. Dalio suggests the “small world” is too small for this ideological battle. I’d go one step further to suggest that, with slowing population growth globally, that’s a huge loss of cylinders for the economic growth engine, which means more fighting over less growth. Even more provocative, however, was this opinion piece in the NYT suggesting there is no actual ideological war – instead, the West and the East are equally bad surveillance states. Extrapolating further, one could say: in China they call it communism, surveillance, and re-education, while in the US and Europe we call it Google, Facebook, Instagram, and Amazon. The point being that, in general, humans have given up their privacy (and, to some degree, their autonomy) with the advent of the Information Age – in the US it seems to be perhaps a little more by choice and convenience. In terms of investing during the Trade War Era, given the increased uncertainty and broader range of outcomes, my suggestion is to bake in a higher structural cost of cross-border business with China, and expect that some technologies might simply be banned with great consequences.
The “Long Term Stock Exchange” (LTSE), strikes me as a misguided effort. Backed by Silicon Valley heavyweights like Marc Andreessen, LTSE won approval from the SEC this week. I addressed this a while back, but it’s worth revisiting, especially in light of the public market reaction to the Lyft and Uber IPOs this year. There is absolutely zero need for a separate stock exchange to match up long-term investors and long-term-focused companies. Public companies, even newly IPO’d stocks, can heavily self select their investors by the way they execute and communicate on the existing stock exchanges without introducing increase costs or complexities. The existing markets today are excellent long-term judges of value, and it’s very easy for long-term investors to find like-minded companies and vice versa. There is zero information value in short term stock prices or stock movements. Let me repeat that with emphasis: there is ZERO information value in short term stock prices or stock movements. I suspect a lot of this LTSE effort stems from VCs like Ben Horowitz’s having a bad experience with public markets when the dotcom bubble popped. Investors cannot short a company out of business, only a company with a bad long-term business model can put itself out of business. The economics of your business are what they are – math is math over the long term. The negative market reactions to the Uber and Lyft IPO are complex, and, in part, are related to the circumstances of both companies going public later than they could or should have, and both companies still vetting their unit economics now under broader investor scrutiny. That’s the reality for high-growth companies: take it or leave it, but playing the same game on a different “long term” field doesn’t change anything. The drops in Uber and Lyft stock this past week have nothing to do with the long-term network effects and unit economics of the ride-sharing business. Lyft and Uber are transforming the transportation market, a market that will undergo even further transformation over the next decade. There is a wide range of outcomes for both companies combined with significant positive Optionality for the industry. The value of the shares long-term is informed by what you believe the future cash flows will be, not by what the stock price is trading at today. The last point I will make is so obvious that it’s surprising it needs to be made: there are plenty of examples of the stock market getting behind companies that invest heavily for the long term – look no further than Amazon for the poster child.
Speaking of the LTSE, maybe it already exists in Australia: SF-based Life360 had their IPO on the Australia stock exchange in part to avoid the chaos of the US markets, but also in part due to the supercharged saving/retirement/investing mandates in the country that provide for a potentially healthier long-term access to capital – part crazy and part savvy, I think.
Passive funds will overtake active funds in total assets this month. If we want to address the elephant in the room with public markets that wouldn’t be fixed with an LTSE, we should look closely at what will happen with voting power as ownership is concentrated in the small handful of giant passive investing platforms. One of the benefits touted by the LTSE is shareholder voting weighted by length of ownership - passive funds which simply own an index are the longest term shareholders out there, so the LTSE would likely weight even more to passive investors - is that good?
A Chinese state-owned company denied Amnesty International a lease in their NYC building:
“Amnesty International U.S.A. said it was told the organization was “not the best tenant” for a building owned by a Chinese state-owned enterprise.”
World’s smallest violin plays for BlueMountain Capital as they exit the equity business a few months after lecturing PG&E directors that their sole duty was to equity shareholders, not their employees, customers, or the environment.
Some kind coverage this week from City Wire.
Disclaimers:
Nothing in this newsletter should be construed as investment advice. Nothing contained in this newsletter is an offer to sell or solicit any investment services or securities. The content of this newsletter is my personal opinion and may not reflect the opinion of NZS Capital, LLC. I may own long or short positions in stocks discussed in this newsletter. This newsletter is simply an informal gathering of topics I’ve recently read and thought about. It generally covers topics related to the digitization of the global economy, technology and innovation, macro and geopolitics, as well as scientific progress, especially in the fields of cosmology and the brain. I will frequently state things in the newsletter that contradict my own views in order to be provocative. I may change my opinions without updating them in the newsletter. Lastly, I often try to make jokes, and they aren’t very funny – sorry.