I have fallen down the abstraction rabbit hole this week. Abstraction is a word I use often in widely imprecise ways. After Audrey called me out, stating that I don’t think you are using that word correctly, I wanted to understand further. I love words. I love rabbit holes. I enjoy the confidence of being precise, and I hate it when I am using words in imprecise ways.
To help guide me I will begin at an intersection of a handful of words I feel are involved, beginning with the definition for abstraction, which is:
The quality of dealing with ideas rather than events
A good rabbit hole always starts with the etymology of the word abstraction:
Abstraction (from the Latin abs, meaning away from and trahere , meaning to draw) is the process of taking away or removing characteristics from something in order to reduce it to a set of essential characteristics.
Ok, this is the history of where my usage of the word went off the rails, as I began worshiping the computer science gawds, which is closer to how I have been wielding the word.
In software engineering and computer science, abstraction is the process of generalizing concrete details, such as attributes, away from the study of objects and systems to focus attention on details of greater importance.
Now, let’s do a “diff” on those two definitions to try and understand the abstraction us technologists have introduced to the word abstraction.
From “taking away or removing characteristics from something in order to reduce it to a set of essential characteristics” to “generalizing concrete details, such as attributes, away from the study of objects and systems to focus attention on details of greater importance”.
OK, we are moving away from things being essential to of greater importance. The devil in these details I think comes down to who is in charge of the abstraction, and specifically that it is primarily young white men being directed by older white men. But, I think we need to go further down the rabbit hole to understand the details, as well as the devils.
I took a stroll around the abstraction definition neighborhood to better understand the distortion that I may have been succumbing to:
- Business Abstraction - Business abstraction is the process of simplifying complex business information into a more concise and understandable form. It can be used to model business processes, write business proposals, and more.
- Capitalist Abstraction - Capitalist abstraction is the idea that capitalism is driven by a high-level logic that separates concrete realities from formalized operations. It can also refer to the abstract rules that govern capitalist society, such as the value of money and the exploitation of labor.
- Financial Abstraction - Financial abstraction is the phenomenon where currency becomes less concrete, more abstract. Examples of this are the shift from cash, which is very concrete, to debit, credit, electronic transfers (like Venmo and PayPal), and even to cryptocurrency, with the degree of abstractness increasing with each example.
This gives me a little bit more context of why my abstraction of abstraction is the source of my imprecise use of abstraction, but there is another word I feel needs to layer on to understand how it fits in — derivatives. This was another word I have been imprecisely using that I wanted to build up a bit more with two distinct definitions.
- Financial Derivatives - Derivatives are financial contracts, set between two or more parties, that derive their value from an underlying asset, a group of assets, or a benchmark. A derivative can trade on an exchange or over the counter. Prices for derivatives derive from fluctuations in the prices of underlying assets.
- Mathematical Derivative - In mathematics, the derivative is a fundamental tool that quantifies the sensitivity to change of a function’s output with respect to its input. The derivative of a function of a single variable at a chosen input value, when it exists, is the slope of the tangent line to the graph of the function at that point. The tangent line is the best linear approximation of the function near that input value. For this reason, the derivative is often described as the instantaneous rate of change, the ratio of the instantaneous change in the dependent variable to that of the independent variable. The process of finding a derivative is called differentiation.
This quick peek down the derivative rabbit role helps me dial in my usage a little bit better, equipping me with more of what I need to argue why the relationship between API producer and API consumer can be considered a derivative, which I think sets the tone for shift and value of abstractions in a world where everything is being reduced to a transaction (ooh I need to add transaction to my list).
This relationship between the API producer and consumer (aka derivative contract parties) sets that stage for the introduction of another word into this intersection—extraction. Which had the original meaning that I intended, but then once again gets quickly consumed by the digital definitions of data, information, and knowledge—which seems on point.
- Natural Resource Extraction - Resource extraction involves any activity that withdraws resources from nature. This can range in scale from the traditional use of preindustrial societies to global industry. Extractive industries are, along with agriculture, the basis of the primary sector of the economy. Extraction produces raw material, which is then processed to add value. Examples of extractive industries are hunting, trapping, mining, oil and gas drilling, and forestry.
- Data Extraction - Data extraction is the act or process of retrieving data out of (usually unstructured or poorly structured) data sources for further data processing or data storage (data migration). The import into the intermediate extracting system is thus usually followed by data transformation and possibly the addition of metadata prior to export to another stage in the data workflow.
- Information Extraction - The task of automatically extracting structured information from unstructured and/or semi-structured machine-readable documents and other electronically represented sources. Typically, this involves processing human language texts by means of natural language processing (NLP).[1] Recent activities in multimedia document processing like automatic annotation and content extraction out of images/audio/video/documents could be seen as information extraction.
- Knowledge Extraction - Knowledge extraction is the creation of knowledge from structured (relational databases, XML) and unstructured (text, documents, images) sources. The resulting knowledge needs to be in a machine-readable and machine-interpretable format and must represent knowledge in a manner that facilitates inferencing.
Ok. Enough with the definitions, let’s bring things home.
This intersection of abstraction, derivatives, and extraction of digital resources and assets from corporate entities and individuals is what is flowing around my head as I look out across the API space each day. The trick here is, how do I get this down in a single story that might remotely make sense to someone else. It will take several attempts, but here we go.
When I use abstraction, I am wielding it in a way that points to nefarious intent or at the very least deliberate ignorance or negligence by the person doing the abstraction, and that in a world that is being perpetually digitally financialized, there is all of this unregulated deviate trading on top of all of this already worrisome extraction leaking out of these digital abstractions.
I am looking to understand all of this to 1) help me understand where I should be headed career wise, and 2) how to push back on the AI stuff which I see as purely being about power, labor, and financialization of every aspect of our lives using the Internet and now artificial intelligence. Let’s begin with the abstraction of a single event that regularly occurs in our lives — taking a photo.
The abstraction of a photo is something that has been happening for sometime, but I’d say the intersection of Steve Jobs interests in abstracting away this event in our into something digital and the interests of Mark Zuckerberg when it comes to abstracting away this into something even more digital is a damn good place to start on this journey.
Let’s abstract away Steve Jobs and Mark Zuckerberg and just say iPhone and Meta. We wouldn’t want there to be any accountability for actual human beings in what I am about to lay out. ;-)
I would say the abstraction occurring through an iPhone, a “photo event” and abstracting that event into a digital image is a fairly precise use of the word abstraction. I think where we build on that abstraction using the Facebook Media API is where the imprecision of the word I use begins to come into focus (pun intended) for me.
IPhone’s abstraction is closer to the actual definition of the word and is “dealing with ideas rather than events” and “taking away or removing characteristics from something in order to reduce it to a set of essential characteristics.” Which maybe I am off the mark, but feels like not just abstraction, but digital abstraction. Which is an important part of the reduction that I want to get at, but this situation is the less nefarious type of smoke and mirrors abstraction I am looking for. Still room for abuse here, but time will tell.
Meta’s abstraction of the photo event begins to move more into my territory in which the abstraction that occurs through the delivery of the Meta Media API made available as a digital resource by Meta. Meta uses this digital resource in their web and mobile applications to publish photos that are uploaded or taken via an iPhone or other device.
Meta’s abstraction shifts us closer to this computer science definition of abstraction, which is the “process of generalizing concrete details, such as attributes, away from the study of objects and systems to focus attention on details of greater importance.“ The design of Meta’s Media API is what is codifying and applying the abstraction at scale to millions of photos. It is what takes the digital object and decides which attributes are of greater importance based upon the Meta system, as well as the context of the user in which the image is being uploaded for, but also for any 3rd-party developer or partner who is accessing the photo via the Meta API.
Meta has a contract around a single media resource in which it has entered into a financial agreement with many partners and 3rd-party developers — this is an API derivative. A single API resource (media) defined as an API contract that is made available to partners and 3rd-party developers. Now, the value derived from the API contract between Meta and API consumers (including their own 1st-party applications) is reconciled as part of a separate Meta Advertising API derivative in which the actual financial speculation and betting occurs—and where the “fluctuations in the prices of underlying assets” occurs—aka market.
This is where I believe the switch gets flipped between Meta providing an honest abstraction of a digital resource to one that becomes extractive—at the point where a digital resource is being abstracted away and turned into a derivative. This is what I have for years called, “reducing everything to a transaction”, and something that occurs across hundreds of Meta digital resources, and across thousands of other resources offered by tech companies.
This is the first time I’ve ever been coherently (hopefully) been able to tell this story, and get out of my head.
Everything changes when the digital resources we depend on in our personal lives get financialized and turned into derivatives. The incentives change. A market is created on our digital resources like images, messages, and payments. Our very human existence gets abstracted away into hundreds of little financial deviates that are betted and traded upon via advertising networks.
Additionally there are hundreds of other applications all targeting us for extraction, and there are secondary markets being built right now to bet, speculate, and trade on these companies behind that extraction. All of these companies are trying to dethrone Meta in the extraction of value from users and running the market where everyone bets, speculates, and trades. This is what is happening with AI, just at an entirely new scale and scope. However, I think I will stop here.
This has given me a framework to think about other API resources like Stripe and Twilio, and the role they play in defining the API economy. This provides me with a more precise understanding of why my understanding of the definition of abstraction was off. This provides me with a more precise usage of the word derivative when I apply to APIs. It also helped me pinpoint the reason good APIs go bad when they become successful derivatives or are bundled into other more complex API derivatives. The layers of abstraction make me dizzy, and trying to understand the honesty of the intent at each layer always feels like a circus funhouse of mirrors.
Where all of this really becomes problematic, is who is in charge of the abstraction. Who gets to decide which attributes go away and which attributes are essential or important. I’d say this becomes much more effective when you have young white men, or increasingly just young men developing the resources and building the abstractions based upon the requirements of older mostly white men. What is essential and important to these older white men, and is codified by these young men, is where the problem lies.
Additionally, what is deemed essential and important in digital abstractions is defined by the contract parties present in each layer of abstraction as defined by derivatives, as well as the protection from observability and accountability that exists at each layer of abstraction (API, company, industry, geographic). There is a lot of protection afforded to those who know how to navigate the layers of digital abstractions, profiting from the extraction, and producing new and exciting derivatives to bet against. Kind of like how mainstream financial markets have worked for sometime but digitally obfuscated, but operating like crypto markets but on our personal lives, social interactions, and feeding off our communities or the increasing lack of communities.
I think the last wave of information fueled financialization had only a handful of folks who were profiting off market volatility, and these will be the ones who exclusively profit off the next generation of unregulated financialization of our personal lives, social, and culture, where the default state is unregulated chaos and volatility.
