Welcome to Stuff I Thought About Last Week, a collection of topics on tech, innovation, science, the digital economic transition, the finance industry, the vagus nerve, and whatever else made me think last week. Please grab me on Twitter with any thoughts or feedback.

Click HERE to SIGN UP for SITALWeek’s Sunday EMAIL (please note some ad blocking software may disrupt the sign up form; if you have any issues or questions please email sitalweek@nzscapital.com)

In today’s post: Google is serious about ecommerce; retail’s linger problem; Cameo, but for everything; Semicon West and DAC recaps; Intel and Taiwan; Vonnegut on society and space exploration; can shortages; rare earths; and lots more below...

Stuff about Innovation and Technology
Vagus Stimulation Quiets Inflammation
The gammaCore Sapphire CV device – a handheld vagus nerve stimulator – has been granted emergency FDA approval for use in COVID-19 treatment. We covered the vagus nerve back in January in SITALWeek #226: “Vagus Nerve is Endogenous Chill Pill: Your vagus nerve connects your enteric nervous system to your brain. In other words, it connects your ‘gut feelings’ to your higher-order cognition. The connection is particularly important for maintaining homeostasis, mood, and the relationship between bacteria and our digestive system (‘The Strange Order of Things’ Damasio p. 133-136). This article on the vagus nerve discusses its ubiquitous importance, including in mood regulation (perhaps because 95% of the body’s serotonin is produced in the enteric nervous system and connected to the brain via the vagus). You can take care of your vagus with stretching, deep breathing, yoga, massage, and other forms of movement.” In terms of COVID-19, stimulation of the vagus nerve helps calm the lungs and may help combat the dangerous “cytokine storm” hyperactive immune response. The tool sends 25 pulses per second at a frequency of 5000 Hz. 

Virtual Fan-Filled Stadium
Microsoft Teams is supersizing its Together mode for video conferencing with 17-foot tall LED screens wrapped around NBA basketball courts so that floating heads of home-bound fans can appear in virtual seats, filling the void where live fans used to exist. It's as weird as it sounds, but with a little imagination and some advanced VR, this could evolve into the main way live sports are experienced in the future.

AI Cracks Fluid Dynamics
Google’s Deepmind released a new model for simulating complex interactions, like those between molecules of water or particles of sand, called the Graph Network-based Simulator (GNS). GNS trains on real-world physical systems, learning through trial and error to generalize to previously unseen circumstances. The video showing the GNS predictions beside the real-world representations is impressive. Simulating fluid dynamics has traditionally been very hard, but AI could open up a whole new set of capabilities. 

Google Ups Ecommerce Ante
Google is finally making a serious effort to operate a competitive marketplace for digital commerce. The search giant is copying Alibaba’s zero-commission model, with the likely goal of making money on ads for products listed. If you simplify the vastly overcomplicated fee models of marketplaces (and include credit card processing fees) you could generally assume that online merchants pay anywhere from around 8% on the low end (Etsy) to 15% on average for Amazon (eBay is usually a couple percent cheaper than Amazon). ‘Buy on Google’ is also integrating PayPal (previously, only Google Pay was allowed) and Shopify, the latter for payments as well as inventory population into Google Shopping. Further, Google has made it so merchants can upload their Amazon-formatted inventory feeds directly into Google Shopping. In other words, it’s for real this time. It’s a no brainer for any multi-platform merchant to put everything on Google and start ringing the cash register with no commission fees. Amazon is now the largest advertiser in the world, spending an estimated $11B per year, with a good chunk (I suspect well over half) spent on Google search ads, so I guess Amazon should become a direct merchant on Google Shopping now and save that money? There is good coverage of the Google marketplace developments on a recent Jason and Scot Show podcast if you’re looking for more detail.

Build to Fulfill, not Linger
Retailers seem to be pinning hope on their ability to serve digital orders from existing stores; however, that strategy has virtually zero chance of success. Stores are built to get you to spend more time and pass more shelves than you want to – it’s why the pharmacy counter is at the back of the drugstore and why it’s no coincidence that the two items you really need are in opposite corners of the store. Very few retail blueprints are optimized for speed and efficiency. Retail consultant Jason Goldberg explains: “Prior to COVID-19, almost all retailers lived by the same two goals: traffic and dwell time. Get as many people to visit your store as possible and encourage them to stay in your store as long as possible. The more people that came to your store and the longer they stayed, the more they would spend, or so the logic went.” When fulfilling a digital order, you want the opposite: the picker(s) to make as few trips as fast as possible. My section last week on the margin problem that grocery stores face in the shift to click & collect/delivery was a popular topic, and it’s just one example of the backwards idea of fulfilling a 2020’s digital order from a 1950’s store design. The CEO of US grocery chain Kroger said in the FT: “When a customer first switches to online, it typically takes three or four years before that customer’s profitability is the same as when they shop in the store.” A better model would be automated, micro-fulfillment centers with vertically-integrated delivery based on routing and subscriptions. Amazon is poised to do well, but for some reason continues to struggle with even the basics of doing online grocery right. Amazon could take 80% of each Whole Foods footprint and remodel it into digital order fulfillment and then only allow customers in for order pickup and to shop the prepared foods section. 

Malls Buying Tenants
Malls are perhaps the retail destination most problematically designed to encourage lingering and eschew efficiency, and a spate of bankruptcies has caused mall operator Simon Properties Group (SPG) to buy retail brands directly. Most recently, SPG is bidding to take over bankrupt Brooks Brothers. Previously, they acquired Aéropostale and Forever 21, and they are working to acquire Lucky Brand as well. SPG’s partner in these deals, Authentic Brands Group, also owns Nautica, Barneys, and Nine West. Other mall retailers that have filed for bankruptcy include J. Crew, New York & Co., and Ann Taylor, as well as department stores JC Penney and Neiman Marcus. The mall operators are in a hard spot because loss of a certain amount of occupancy or anchor tenants often allows the remaining tenants to renegotiate their rent. Therefore, SPG buying failed retail brands is like treating the symptom while the patient is still rapidly dying from the cause.

Hulu Best Positioned to Re-bundle
The lackluster launches of the new streaming apps HBO Max and Peacock underscore the need for a digital bundler to aggregate content for consumers. Despite their large libraries and content budgets, WarnerMedia and NBCUniversal can’t achieve subscriber numbers like Netflix’s 180M or Disney+’s 50M. Despite their streaming head start over Peacock and HBO Max, CBS All Access and Showtime, combined, only had 11M subscribers at year’s end. The politics are complicated and well known, but Hulu appears to be the best option that I see to accomplish re-bundling for consumers. Hulu’s owner Disney is a rival but still more of an ally than the adversarial hardware distribution platforms – Roku, Amazon, and Apple – which falsely view themselves as the winners in the power struggle between content and distribution. Even the traditional cable companies are now friendlier to content creators than Roku and Amazon. Efforts to recreate live TV by the virtual MVPDs, such as Sling TV, YouTube TV, Hulu Live, and AT&T TV Now, are failing because they are still channel centric instead of content or brand centric. A great example of what works in streaming is the new FX on Hulu experience; as FX CEO and master content creator John Landgraf explains in the Variety article on the death of cable channels: “It allows us to maintain — even increase a little bit — our investment in our programming for our linear channels…but where all the growth from investment in the television industry is, is in streaming.” A bundled, digital combination of premium content from the best studios and brands would give me little reason to ever turn on Netflix to watch Unsolved Mysteries reruns. I can’t leave this paragraph without throwing in the obligatory “and live sports”...Afterall, there’s premium content, and then there’s live sports.

Experts On Demand?
Cameo is a platform that connects regular people with celebrities. For a sum of money, you can put in a short request and get back a video message, like wishing someone a happy birthday. I was recently reading about Cameo in Rob Litterst’s newsletter, and this quote from the Cameo co-founder stood out: “What we’re really building at Cameo is the marketplace where, for X amount of money, you can do Y activity with Z person. We’ve just been focused on doing one Y activity with as many people as possible.” This brings to mind the expert networks Wall Street uses to gain insight into industries and companies. Is anyone starting a Cameo for the plethora of industry experts to pop on a paid Zoom for 20 minutes at a time? Of course, the service wouldn't need to be just for investors, e.g., how about having access to a hot tub repair person or plumber to walk you through a quick fix? Or having a decorator on-call to help you arrange that pile of throw pillows for maximum feng shui?

Moore's Law Goes Vertical
NZS Capital’s Jon Bathgate brings us a dispatch from the virtual Semicon West and Design Automation Conference (DAC) last week: One consistent theme across both shows is that Moore's Law is likely going vertical, as we are close to having exhausted our ability to continue to shrink transistors in two dimensions. DRAM and logic devices are likely going to start scaling vertically over the next decade, which will make them look more like NAND, which started scaling vertically in the middle of last decade. Advanced packaging will also play an important role moving forward, where we are seeing chips stacked on top of one another in 2.5D/3D device structures or in "chiplet" architectures in which different parts of the device can be manufactured on different processes and integrated into a package. Stacked silicon architectures have been crucial in AI and other intensive computing applications, as this approach allows the memory to sit as close to the compute engine as possible. TSMC has emerged as an early leader in packaging technologies and has partnered closely with companies like Nvidia, Xilinx, and Apple to drive system performance with approaches like through-silicon-via (TSV) and integrated fan out (INFO). They also discussed another system on an integrated chip (SoIC) last week as a potential next-generation approach. This system-level innovation leans heavily on the ecosystem, with design software makers Cadence and Synopsys playing a crucial role in device design and layout, and etch equipment providers, like Lam Research, facilitating their manufacturing.
 
Intel Calls Pinch Hitter for 7nm
Jon also weighs in on Intel’s further delays of 7nm manufacturing following years of delays with their 10nm process. Intel's server CPU roadmap will now be a full node behind TSMC's (if it ramps on schedule), which puts Intel at an unprecedented process-technology disadvantage to AMD and ARM-based server solutions, like Amazon's Graviton. With this latest in a chain of Intel stumbles, it’s increasingly clear that they will have to outsource more of their manufacturing to 3rd-party foundries, which CEO Bob Swan characterized as a prepped contingency on last week's call. The company has utilized a chiplet architecture for many of their key products, which will allow them to internally manufacture some pieces of the chip while farming out others to TSMC (or Samsung). The first example will be Intel's 7nm GPU: originally intended to be in-house, Intel declared last week that it will have key, leading-edge pieces fabbed externally (likely one reason TSMC is not worried about losing the Huawei business due to US government sanctions later this year). Moving towards a heavier reliance on outsourcing is a major change for Intel after decades of religiously sticking to their vertically-integrated manufacturing model. The decision will certainly have long-term margin implications, but it could solve other problems – like closing the process-technology gap with the competition and easing the burden of their current ~$15B annual manufacturing CapEx spend. As we’ve covered before, the increasing dominance of TSMC represents a fragility for Western companies and is likely to exacerbate already tense relations between the US (who believes Taiwan is its own nation) and China (who does not). The move also highlights the dearth of leading-edge manufacturing capability/capacity on US soil, which, if started today, would take a decade to fully implement.

Miscellaneous Stuff
Beverage Market Sees Can Shortage
The US may be headed for a short-term shortage of 12oz aluminum cans as consumption for fizzy beverages (both alcoholic and non) shifts from restaurants to home and people stockpile for the pandemic. This may prove short term; but, over the next decade, several hundred billion single-use plastic bottles are likely to shift to aluminum, which will require an industry used to growing 1-2% in a good year to grow at 10x that pace. New can manufacturing capacity will need to come online, and long-term supply contracts will be necessary to fund the expansion. Enough aluminum will need to be secured as well. Last week, Polish can maker Canpack announced a $366M investment to establish their first US plant, in Lackawanna County, Pennsylvania, home of Canpack’s parent company, mushroom grower Giorgi Holdings.

Vintage Vonnegut 
This 1973 interview with Kurt Vonnegut is so rich in insight that I struggle to choose the best quotes to pull out of it, but here are a few:

• “And everything is a lie, because our brains are two-bit computers, and we can’t get very high-grade truths out of them. But as far as improving the human condition goes, our minds are certainly up to that. That’s what they were designed to do. And we do have the freedom to make up comforting lies. But we don’t do enough of it.”

• “Human beings will be happier—not when they cure cancer or get to Mars or eliminate racial prejudice or flush Lake Erie but when they find ways to inhabit primitive communities again.”

• “What I say didactically in the introduction to Breakfast of Champions is that I can’t live without a culture anymore, that I realize I don’t have one. What passes for a culture in my head is really a bunch of commercials, and this is intolerable. It may be impossible to live without a culture.”

• And, this quote, which I’ve often considered to be the motto of SITALWeek: “You understand, of course, that everything I say is horseshit.”

There is a common thread in much of Vonnegut’s writings that’s evident in this interview: our need to have a common culture and our uniquely human ability to completely invent what that culture is. No culture is rational; so, just pick the best attributes you want, and the brain will play right along. The Internet could have – and should have – given us a global sense of shared identity, but it’s thus far done the opposite. Which reminds me of another Vonnegut quote: “We are what we pretend to be, so we must be careful about what we pretend to be.” Lastly, you might get a new perspective on Elon Musk and SpaceX after you read the interview...🚀

Space Exploration’s Hotter than a Rocket on Re-entry
China launched the second of three global missions to Mars this summer. The Tianwen-1 mission has 13 scientific instruments, including tools to study the atmosphere and magnetic fields, a rover, and an orbiter at its disposal when it reaches the red planet in February 2021. The Hope orbiter, launched by the United Arab Emirates last week, constituted the first mission (and first ever for the UAE), and NASA’s Perseverance rover is scheduled to launch this coming week. In other space exploration news, Technology Review reports that Japan has officially teamed up with the US for NASA’s project Artemis to put “boots on the moon” in 2024. 

Humans Afoot in North America Much Earlier than Thought
One of the plot arcs of archeology over the last couple of decades has been that humans were movers and shakers long before we thought. We left Africa earlier, said “how ya doin?” to Neanderthals much earlier, and ventured to North America much earlier. Now, new evidence suggests humans traveled to the Americas much, much earlier than previously thought. A study published in Nature puts humans in North America before it was last maximally covered in sheets of ice. That would mean closer to 30,000+ years ago instead of just ~15,000 years ago. For example, thousands of stone tools found in a cave in Mexico (at an elevation of 9,000 feet above sea level) provided a convincing chronology, with the oldest dating to 31,500 years ago, suggesting humans set up camp on the continent as early as 33,000 years ago.

Stuff about Geopolitics, Economics, and the Finance Industry
Reshoring REE 
Japan and the US (along with Australia and Canada) are working to displace China’s stronghold on rare earth elements (REE). The 17 REE are important for the manufacture of just about everything important that plugs in or has a battery these days. The elements themselves aren’t necessarily rare, they are just widely dispersed and infrequently found in large quantities. China currently accounts for 70-80% of global REE mining, and that’s down from 97% a decade ago. Japan has a large supply of REE-rich offshore mud; meanwhile, the US is reopening mines in the Mojave and developing a compound known as Lanmodulin (LanM), discovered by Lawrence Livermore National Laboratory (LLNM). LLNM’s LanM is a protein that binds selectively to REE, allowing extraction in an environmentally friendly way from multiple sources, including recycled electronics.

✌

Disclaimers:

The content of this newsletter is my personal opinion as of the date published and is subject to change without notice and may not reflect the opinion of NZS Capital, LLC.  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. Often I try to make jokes, and they aren’t very funny – sorry. 

I may include links to third-party websites as a convenience, and the inclusion of such links does not imply any endorsement, approval, investigation, verification or monitoring by NZS Capital, LLC. If you choose to visit the linked sites, you do so at your own risk, and you will be subject to such sites' terms of use and privacy policies, over which NZS Capital, LLC has no control. In no event will NZS Capital, LLC be responsible for any information or content within the linked sites or your use of the linked sites.

Nothing in this newsletter should be construed as investment advice. The information contained herein is only as current as of the date indicated and may be superseded by subsequent market events or for other reasons. There is no guarantee that the information supplied is accurate, complete, or timely. Past performance is not a guarantee of future results. 

Investing involves risk, including the possible loss of principal and fluctuation of value. Nothing contained in this newsletter is an offer to sell or solicit any investment services or securities. Initial Public Offerings (IPOs) are highly speculative investments and may be subject to lower liquidity and greater volatility. Special risks associated with IPOs include limited operating history, unseasoned trading, high turnover and non-repeatable performance.