Nihal Uppugunduri's WebsiteMachine learning today is a major subfield of artificial intelligence, and a field that has significant applications in other subfields of AI and tech in general, such as in natural language processing and computer vision. In this post, I hold that the definition of this subject, at least as it is given today, is fundamentally informal. There are a couple candidates for definitions that we will consider.
Edit from the future (2025): This post’s points are now outdated according to my current thoughts, given technological developments since.
Artificial intelligence, or AI, is a loaded term. When it was first used, it would refer to the abilities of computers to play perfect-information games like chess. But now that is “old tech,” and we don’t consider that as much a part of AI. In fact, I contend that AI has always been the term used to refer to the bleeding edge of computer science at any given moment. It’s why we might never “solve” or “figure out” AI, since once we figure out AI today, the term AI will just be redefined tomorrow to be the new horizon to reach for.
Nevertheless, AI has taken on a distinct character recently that might prompt us to consider a closer definition of AI in terms of what it means today. This is the question we will investigate now.
When we learn math, a typical approach is the following. We discuss some motivation, based on what we have already done. This motivation leads, in a possibly informal way, to a formal definition. (It could also be a set of formal axioms, but here we encompass that, as well as any other type of formal “framing,” under the term “definition.”) We then work with our definition using formal logic to produce theorems.
This can be a great approach. However, there are important caveats that we must keep in mind, which I call attention to in this post.
Assume we are developing the curriculum for an extracurricular math program for an advanced student. How could we introduce formal reasoning and the axiomatic method, two cornerstones of the practice of math today, as early as possible? And specifically, what are the implications for this from the philosophical discussion of the formal foundations of mathematics, in my post On the Circularity of Mathematics?
It is a well-known, but uneasy, admission that the foundations of mathematics seem to be inherently circular. Indeed, we may for example list out the axioms for the positive integers, say using the Peano formulation, but in writing down this list we already use numbers: identifying the first axiom, then the second axiom, and so on. Even if we argue that we need only formally specify a set of axioms rather than a list, this still makes the axiomatic formulation of sets circular, whether with Zermelo-Fraenkel or another system.
You’re patting yourself on the back, proud of the code you’ve just written for a cool new feature. You’re itching to add it to your organization’s repository and see customers use it as soon as possible, but you can’t. You now need to write a bunch of unit tests that meet your organization-mandated “code coverage” level. You begrudgingly sit down and start to hammer out the tests, thinking, why? “I’ve already written the code, why do I have to go through this bureaucracy?”
In this post, I’ll discuss what unit tests are useful for, why they can seem so annoying to write, and how to write them more effectively.
Before last year, online education was somewhat more of a Silicon Valley experiment in the ways tech could enhance aspects of education and address some common issues with it. With the advent of COVID-19, schools that ordinarily wouldn’t have considered such a radical proposal anytime soon have been forced to live and breathe it fully. It’s not a stretch, then, to say that the kinks are still being ironed out in delivering effective online education. But as it becomes more prevalent and important, we must address those issues now more than ever. Let’s talk about one of the long-standing arguments against online courses that has been continually invoked until recently: classroom interactivity.
Google. Apple. Amazon. Facebook. Microsoft. They all market directly to consumers. However, their consumer products, especially when free of cost, can be supported by high-value business models for enterprises. For example, AWS is a major cash cow for Amazon, and in general, businesses pay big money for enterprise tech for later payoffs due to the increased efficiency. Additionally, innovation in enterprise tech has a positive developmental feedback effect on consumer tech, whereby production processes can be improved and form compounding sequences of “improving improvements”. Enterprise tech thus produces fundamental value for enterprises and consumers alike.
