เกี่ยวกับตอนนี้
Good evening.
As readers of this newsletter know, I am skeptical of crypto’s role as a currency, but remain curious as to its potential as an asset. In my January 10 update, I wrote:
[A]pplying the innovator’s dilemma framework to our subject matter, if crypto is the disruptor, who are the incumbents? There would seem to be two. First, government-issued fiat currencies fulfill the role of money as we traditionally understand it. Second, there is gold, or rather, any alternative asset class that provides diversification.
In the first instance, it’s not a fringe act today to note the problems of fiat currency. Monetary policy-induced overleveraging and rising geopolitical risks are just two examples most frequently mentioned in recent news. But these shortcomings just prove that, unlike the innovator’s dilemma, government fiat currencies are far from “over-serving” the market. On the other hand, crypto, atypical of a disruptive technology candidate, is not simpler nor more convenient to use. Perhaps no one makes this intuitive point more poignantly than the comedian John Oliver, when he called bitcoin “everything you don’t understand about money combined with everything you don’t understand about computers.”
The second instance of crypto as an alternative asset class is more promising. With the stock market at all time highs yet still flush with cash, in recent weeks investors have directed significant sums into crypto…Coincidentally while crypto was breaking price records daily, we saw a cratering in the precious metals market last Friday…Relative to precious metals, which require familiarity with financial securities and fintech, crypto still may not be simpler, but it may not be much harder…
The following week at the virtual World Economic Forum, Enass Abo-Hamed of the Royal Academy of Engineering, Imperial College of London and Daniel Evans of the Gibraltar Stock Exchange make a novel argument that cryptocurrency is a battery.
The value proposition of crypto mining as storage of energy, or rather, of value, flips the carbon footprint argument of crypto on its head. It introduces an element I haven’t thought of previously, which is that excess renewable energy must be stored or used somewhere when the grid is overloaded. Rather than turning off a wind turbine or solar power field, the excess energy is converted via crypto mining into a storage of value that could be exchanged for services later: in other words, a currency. Is this enough to tilt the scale in crypto’s favor as a currency according to the innovator’s dilemma framework? Not for me, because to disrupt fiat currency, crypto still has to get much easier for the uninitiated user to learn. But does this reframe the case for crypto, especially as a counterpoint to critics who have argued crypto is bad for climate change? Absolutely.
This brings me to a more general question. As our society pushes forward on new technologies, are we creating more sustainability problems than we are solving? Take for example artificial intelligence. Much has been covered in the expanding use cases of AI/ML, but what about the economics and environmental footprint?
To unpack the problem, let’s look at what AI/ML entails. First, at the physical layer, more objects in our lives are embedded with sensors or are sensors. The aim and consequence are to record and quantify more of our physical world, creating vast data sets that make machine learning possible. Second, at the application layer, growing compute power and model sophistication such as dimensionality reduction and manifold learning have helped data scientists tackle messier technical problems than could be addressed even just a few years ago.
Meanwhile, the pandemic has heightened the urgency and societal stakes to direct research insights toward practical problem-solving, broadly defined. In the words of famed Silicon Valley entrepreneur and investor Marc Andreessen, “It’s time to build.”
And yet as practitioners know well, AI/ML faces particularly challenging economics because of data and model complexities.
An underlying cause is entropy, the tendency of our natural world to become more chaotic over time. In my January 24 explainer of David Christian’s Big History theory, I wrote:
Think of your desk. No matter how much you tidied it yesterday, unless you just cleaned up, it’s likely to now be a mess. The same applies to star systems and intergalactic space. This tendency towards chaos is also known as entropy.
Human progress, then, could be understood as our activities to reduce entropy. We make progress when energy is more efficiently converted to work, and when the information we gather improves in quality.
Another cause is the long tail: no matter how neatly a model accounts for a perceived cluster of data points, many still fall outside the explanatory power of the model, forcing the data scientist to search for yet more advanced modeling techniques. Financially, these technical challenges impose a high level of recurring expenditures for modeling and training data generation, meaning such costs must be classified as operating expenses rather than capital expenditures.
This is no trivial matter.
Expenses classification affects the shape of financial statements, which directly affects capital budgeting. Without a path toward a sustainable economic model, investments in AI/ML research would be supported by just a handful of the most well-resourced companies. Despite best intentions, management attention at any organization is by necessity constrained, limiting the span of subjects that can be researched, and even then, at potentially great computational, economic, and environmental cost.
Back to the question at hand: what is the role of the tech sector in our effort to create a more sustainable economic growth model globally? There is evidence to support an optimistic view. According to GeSI, a global NGO focused on sustainability in the technology sector, for every ton of CO2 emitted by the technology sector, up to 10 ton could be saved in other sectors in the economy. Similarly, a Frost and Sullivan report showed that connected technologies can take 20 million cars off the road by 2025.
As the World Wildlife Foundation reported recently:
The biggest role that the digital sector can play is in contributing effective solutions to other sectors, influencing consumer and producer behaviour and leading the transformation of our energy systems.
Still, while a lot of these arguments sound reasonable, without a global standard and systematic way of tracking and accounting, it would be too easy for them to remain in principle but not reality. That, perhaps is the greater call for action as we slowly wobble out of the pandemic and try to get in front of the next crisis.
If you’ve been forwarded this update, please consider subscribing below. Thank you for your support.
From Aspen, Colorado 🇺🇸
Victor
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit iondrive.substack.com
