While "general AI," or a general purpose thinking machine, has captured the collective imagination, the extraordinary progress in the past decade has focused on "narrow AI," or problem-specific approaches.
- [Instructor] When we talk about Artificial Intelligence, one important distinction to keep in mind is the difference between General and Narrow AI. Now General AI, or Strong AI, is the idea that AI can create general-purpose thinking machines that can reason and learn on many topics, just like a human, and this is the sort of thing you see in science fiction and the popular imagination. The computer that operates just like a person, only better. And it's true that that was the focus of a lot of early research, back in the 1950s.
However, the amount of information required to do general purpose thinking, and the number of decisions that must be made grow exponentially in what's called a combinatorial explosion. There are so many possibilities, that it very quickly outstrips any computing resources. And so, it was found that General AI really couldn't work. It wasn't feasible. It may be feasible at some point in the future, but because of the explosion of possibilities and resources, researchers had to give up research on that overall goal.
Now, as opposed to General or Strong AI, there is Narrow, or so-called, Weak AI. And this is artificial intelligence that focuses on a specific domain, like categorizing photos, or identifying spam email, or approving credit card applications, and it's trying to solve just that one problem. And what this does is it dramatically reduces the size and the complexity of the problem space. It reduces the number of permutations that must be considered when making decisions. And consequently, most of the growth in AI in the past decade has been in these focused domains.
And that has made it much easier. It's still enormously difficult, but it's made it much easier, and really a bounded problem, and something that could be solved using the resources that are available. And that can explain, in part, the growth in AI over the last decade. AI has passed from being some sort of intellectual curiosity for people stuck a lab, to becoming a business necessity. The AI market is projected to grow between 125% and a 150% each year for several years.
This is amazing growth, and most of it has come in the form of Narrow, or again, so-called Weak AI where people are simply trying to solve very specific problems, and those are the examples that we're going to be using throughout this course.
- Bias in AI
- Navigating the social challenges of AI
- Moral reasoning and relational ethics
- General Data Protection Regulation (GDPR) and AI
- Discrimination in data
- Liability and AI
- AI in life-and-death situations
- Confronting the challenges of AI as a developer, executive, and consumer