SITALWeek #404
Welcome to Stuff I Thought About Last Week, a personal collection of topics on tech, innovation, science, the digital economic transition, the finance industry, and whatever else made me think last week.
Click HERE to SIGN UP for SITALWeek’s Sunday Email.
In today’s post: console gaming might be bottoming out after the pandemic hangover; both gaming and YouTube might benefit from extended Hollywood strikes; the erbium-doped fiber amplifier boom and bust still holds lessons for today's investing; muscling out the competition with AI spending; what kinds of jobs pay you far into the future for work today?; the simple joy of YouTube's walking tours; and, much more below.
Stuff about Innovation and Technology
Lessons from an EDFA
One of the biggest unsung inventions underpinning the global Internet was a device that allowed amplification of optical signals without converting back and forth to electronic signals. The invention – erbium-doped fiber amplifiers, or EDFAs – was first described in 1985 by Robert Mears, then a graduate student at the University of Southampton, and eventually became a billion dollar business. These inline amplifiers allowed for much higher speeds and lower signal loss for long-haul fiber-optic communications, such as undersea cables and the network that comprises the modern-day Internet. I have a fond memory of EDFAs as one of the first areas of research that I could really dig into in the late 1990s where my physics background proved useful in helping me to see the potential market and why it was such a game changer. And, the role of EDFAs in the communications stock crash, which came nearly a year after the dotcom crash, proved to be a foundational lesson in investing booms and busts. I have a hazy recollection of penciling out the problem that comes when you increase efficiency by 4x but only charge 2.5x more, as many of the fiber infrastructure providers were doing at the time. The optical equipment market bubble culminated in a series of mergers between manufacturers of EDFAs and other optical components. JDS Uniphase, after acquiring SDL Inc. and E-TEK for $63B (in inflated stock values), peaked at a market cap of over $120B, only to see its value fall by more than 99%, resulting in the largest merger writedown in history at that time. The telecom spending crash was of course complex, but the groundwork was laid by the rapid deployment of ultra-high throughput cabling and signal boosters, which resulted in a massive glut of capacity. This lesson has stuck with me for a long time: sometimes a technology is so disruptive that it can quickly fulfill its purpose and shrink to an industry of much lower importance. Of course, Internet traffic today is around ten billion times bigger than it was in 2000, but, for the first few years of usage growth (which are barely perceptible in the exponential curve) the infrastructure had already been deployed to satisfy bandwidth needs. As we look at the insatiable demand for chips to train and run today’s AI models, combined with the efficiency gains that AI itself will create in software and chip design, we may want to keep one eye to the past and the other to the future to balance our expectations.
PG&E, We Need More Power!
Speaking at a PG&E event, Elon Musk said utilities need a bigger sense of urgency to prepare for increased electrical demand. “Musk urged the power sector to take a closer look at demand projections amid the energy transition. Global energy consumption today is roughly a third electricity, a third for transport and a third for heating, he said. ‘So even if you assume that electricity demand is static, in order to transition to a sustainable energy future where everything is electric and sustainably electric, we need a tripling of electrical output,’ he said.”
Alphabet’s AI Blitz?
Google recategorized its spending in the newly merged Google Brain+DeepMind AI division in its earnings report last week. The sum now resides in “Corporate Costs, Unallocated”. While this line item does fluctuate a lot due to hedging and restructuring costs, it tended to run ~$300M before jumping to just under $1.2B in the most recent quarter following the allocation change. So, one might hazard a guess that Google is spending between $3-4B annually on AI research and product development (the actual number could be lower or even far higher and would include things like engineering on the company’s TPU chips, etc.). It’s a number that dwarfs even the biggest VC infusions to AI startups and certainly rivals Microsoft/OpenAI. Previously, AI spend was housed in the division that encompassed Google’s lucrative search advertising business. Why break it out now, given that AI is becoming the heart of Google Search? Well, surely DeepMind is broadening its scope beyond search, but one might also speculate that Google’s game theory is to make it easier for people to see how much money they are throwing at AI, which might give VCs and aspirational entrepreneurs a bit of a pause. As the FT notes, seven of the eight key transformer experts at Google (the genesis of today’s LLMs) have left to form AI startups. Of course, spending a lot of money can be meaningless if a company is mired in bureaucracy and paralyzed by Innovator's Dilemma, rendering them incapable of disrupting their lucrative businesses even when it’s the right long-term choice. I have noticed a marked improvement in the (terribly named) Search Generative Experience in Google’s Search Labs in the last couple of weeks as AI increasingly takes over search, and I’ve also noticed that ads are starting to populate the AI search results. This progress perhaps foreshadows the anticipated launch of Google’s ChatGPT-competitor Gemini, which brings together LLMs and DeepMind’s AlphaGo-like neural nets.
Regardless of the sums being spent, Google is moving rapidly across a number of AI projects. Back in March, I discussed PaLM-E (see LLMs Transforming Robotics), Google’s initial attempt at integrating an LLM into a robot. The NYT covered this same effort last week, and DeepMind also has a new post on what they now call RT-2. As I’ve said before, I believe the single biggest disruption over the next decade will come from LLMs embodied in physical form factors.
Adobe’s AI Quandary
Adobe employees have internally discussed their anxiety that the company’s rush into AI tools may be a threat to their graphic designer customer base, according to Business Insider. This is of course the Catch-22 I’ve covered many times concerning our choice to fight or embrace AI to move to higher levels of value add. From the article: “Other messages in the Adobe Slack channel were more critical of the AI revolution, calling it ‘depressing’ and an ‘existential crisis’ for many designers. One person said some artists now feel like they are ‘slaves’ to the AI algorithm, since their jobs will mostly involve just touching-up AI-generated work. Some had a more positive view. Photoshop made artists more productive, and AI will only increase their efficiency, they said. One person said many freelancers and hobbyists will benefit from the increased output, even if some companies reduce their design workforce.”
Dividing by Infinity
In other frontline AI news, Breaking Bad star Bryan Cranston, addressing Disney CEO Bob Iger, said: “We ask you to hear us, and beyond that to listen to us when we tell you we will not be having our jobs taken away and given to robots. We will not have you take away our right to work and earn a decent living. And lastly, and most importantly, we will not allow you to take away our dignity!” I addressed this topic in detail last week, so I won’t bore you with a rehash. But, one new question did come to mind: what are the types of jobs where working today assures you contractual compensation far into the future, like the way Hollywood operates? Surely doing a good job, no matter what your job is, is likely to help you stay employed over time (and some folks are lucky enough to have jobs with pensions). But, how many people make direct money today from a nail they hammered 50 years ago, even if the building is still standing and being used? It’s sort of an odd concept if you think about it, and, yet, I also understand the reasons for it – if your work is still generating new sources of revenue, should it not also generate payments to its creator, like a license on a patent? But, the problem is that, in today’s content-saturated world, the typical production costs and backend payouts of a Hollywood film are way too high given the work’s vastly diminished staying power and value. The DCFs that determine future payments were fed with the wrong assumptions (go figure, of course DCFs never work), and the studios are correct when they argue against striking actors that future residuals will be paltry. Investing is a rare job where, through the magic of compounding, you can continue to get paid year after year based on early good performance. But, these sorts of situations seem unique, and, for the most part, people get paid the current market value of their work and then move on to the next task.
Gaming Bottoms Out
Microsoft noted that its content and services revenues from the Xbox gaming division grew 5% in the most recent quarter. This figure is notable given that the 2020 pandemic surge in console and PC gaming flatlined in 2021 and declined (along with mobile gaming) in 2022. With Hollywood turning into Vaudeville amid strikes and shifting content consumption habits, gaming stands to benefit, particularly a few years from now as generative AI and LLMs transform interactive content. Broadly speaking, our content consumption is now zero sum (see Spiraling Content Meets Maxed-Out Attention from January 2022) – in order for one medium to do better, another has to lose share (until we have 24/7 AR glasses and neural links beaming content into our brains).
Miscellaneous Stuff
Walking Tours
Thanks to a tip from Kevin Kelly’s Recomendo newsletter, I discovered the growing trend of walking videos on YouTube. These meditative journeys are often multi-hour captures of quiet, first-person POV strolls through cities at night in the rain, natural surroundings, or other destinations around the world. I took special note of the walking tours of Disney parks, and the irony that Disney’s $200M blockbuster movies are now competing for our viewing time with simple, low budget videos of someone just walking around Epcot. YouTube, which returned to growth in the June quarter as well, is also likely to be a major beneficiary of any lull in Hollywood content.
Concerning Cancer Trends
STAT reports on the worrying and unexplained rise in younger incidences of cancer: “At age 35, someone born in 1990 will face quadruple the risk of rectal cancer and double the risk of colon cancer when compared with the risk faced by a 35-year-old who was born in 1950.” The article speculates on the areas of lifestyle changes researchers are focused on: “Whether it’s sitting all day, consuming cured meats and sugar-sweetened drinks, taking antibiotics, or staying up late with the lights on, these practices — their effects probably interacting — seem to have had a profound impact on the internal workings of our bodies, disrupting metabolism and boosting inflammation. Starting early in life and accumulating over the years, these behaviors can promote cancer in some people, in ways that are little understood.”
Stuff About Demographics, the Economy, and Investing
Israel’s Tech Tensions
Israel has long been a source of innovation in the technology sector, notably in software and security tools. Small startups are often acquired by major Western tech companies, and many large, standalone businesses have launched onto the global stage, creating significant market value relative to the size of Israel over the last few decades. Intel, for example, acquired (and then spun out) Israeli autonomous vision startup Mobileye, which has a current market value of $30B. Intel has 12,000 employees – 10% of its total workforce – in Israel today (where it’s operated since 1974) and is in the middle of a $25B expansion there. I am by no means an expert on Israeli politics, but this Wired article discusses the changing political situation: “In January, Benjamin Netanyahu—in his sixth term as prime minister and backed by a coalition that includes far-right parties—introduced a bill designed to weaken the powers of the country’s supreme court. Supporters of the plan say it’s needed to prevent the court from intervening in politics. Critics say weakening the reform will erode democracy and hand unchecked power to the government. Despite huge protests, Israel’s lawmakers backed the first part of the judicial overhaul this week.” Wired notes this move is causing many tech startups to relocate outside of Israel (or make plans to do so). It appears concentrating engineering talent in this crossroads country, which has always been somewhat of a risk given the region’s historically fraught geopolitical situation, could be more perilous going forward.
✌️-Brad
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 an informal gathering of topics I’ve recently read and thought about. I will sometimes state things in the newsletter that contradict my own views in order to provoke debate. 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.