Intelligence

By Stephen Downes - Stephen's Web. Why Some AI Efforts Succeed While Many Fail
Irving Wladawsky-Berger, 2020/01/20
I remember when digital technology came into the institution in the 1980s and 1990s the oft-asserted (but less often heeded) advice was that the benefits of technology would only be realized by transforming operations to take advantage of it. Through the decades since this has proven to be true over and over again. And so it's not surprising to read a similar analysis of institutional AI. "Integrating AI and digital transformation initiatives is particularly important, since both typically require large-scale, enterprise-wide efforts to redesign work processes, systems and structures". More...

By Stephen Downes - Stephen's Web. How artificial intelligence will impact K-12 teachers
Jake Bryant, Christine Heitz, Saurabh Sanghvi, Dilip Wagle, McKinsey, 2020/01/16
This article is a mixture of hoary old truisms (eg., "their role will shift from instructor to facilitator and coach") and generic 'technology will make things better' but which nonetheless reassure educators by saying that not only will tech not replace them, the demand for teachers will actually grow in the future. The premise here is that new technologies will reduce time spent on administrative tasks, allowing teachers to spend more time on teaching tasks (or maybe it will cut down on those 50-hour work weeks - either one would be good). But the link between artificial intelligence and these outcomes is pretty tenuous - it involves things like using AI assistance to grade papers and having AIs fill out forms. There's a lot of talk about AI enabling personalization - but the math doesn't work. If tech is saving 30 percent of a 50 hour work week, the it provides an extra 6 minutes per student per day. More...

By Stephen Downes - Stephen's Web. Could exams be marked by AI? Ofqual launches ‘exciting’ competition
Freddie Whittaker, Schools Week, 2020/01/09
I really think it's only a matter of time before all student work is marked by AI. At a certain point, it will be hard to justify using human markers when AIs are demonstrably more fair and more reliable. But of course, this still needs to be demonstrated. That's where this work comes in. More...

By Stephen Downes - Stephen's Web. Adoption of AI in education is accelerating. Massive potential but hurdles remain.
HolonIQ, 2019/04/05
I saw this a few days ago, didn't list it, and then needed it almost right away. That's a pretty good sign I need to list it. This summary describes how artificial intelligence is being used in education. What makes it useful are the breakdowns - differemt AI applications, different uses in education, different reasons for adoption, where AI is creating value, etc. More...

By Stephen Downes - Stephen's Web. Exploring Minds: Modes of Modelling and Simulation in Artificial Intelligence
Hajo Greif, Perspectives on Science, 2020/01/01
Ignoring the title and main purpose of the paper (29 page PDF) for a moment, I encourage people to look at the first section for a really useful discussion of the various types of models that have been proposed through the years. Models might "bear observable and pre-theoretic similarities to their target systems." Or maybe they "are structures that express interpretations of a formal theory." Or maybe they " are structures that bear a formally defined relation of isomorphism to their target system," as Bas van Fraassen suggested in 1980. Whatever your definition (semantic, syntactic, or other) "models are epistemically more central to science than theories."
All of this becomes really important if you think of minds as being (or containing) models of knowledge, the world, or whatever, whether they are representations, reflections or whatever. More...