Is it just me, or has this week been the fastest in all history? I can barely believe I am here again with a whole week’s worth of new AI news to share. Perhaps it has something to do with the fact that I had last week off work for half term so, newly rested, I’ve powered ahead like a racehorse these last six days. It’s been fun and it’s been productive. Let’s hope the coming week is just as fruitful!
Anyway, blah blah, time to keep galloping on with my productivity so I can get on with preparing my Best Woman speech for my bro’s upcoming wedding.
Most Socially-Shared Articles of the Week
Will it or won’t it? It’s been an issue that’s dominated the top AI articles week in week out for at least as long as these reviews have been running. I’m beginning to make my mind up as to which way up the penny is going to fall. Like Kai-Fu Li, the author of this article, I believe that we are in for a shock as we continue to underestimate how much of an effect AI is about to make on the job market. The realisation that we marched ahead without being ready is going to hit us like a slap to the face, and – as is usual for stupid biological humans – we’ll be leaving our kids to pick up the pieces.
Speaking of AI’s effect on the job market, Mark Cuban has also made his mind up. “The hard part isn’t whether or not [AI] will change the nature of the workforce — it will. The question is, over the period of time that it happens, who will be displaced?” Cuban suggests the best course of action for those considering which career path to take might be better off with something less vocational than one that teaches lateral and creative thinking, like philosophy. Cuban, incidentally, also had this to say about artificial intelligence: “It scares the shit out of me”.
AI is going to be huge in Healthcare and, in many ways, it already is. This example of how eye scans using a machine learning algorithm may be a quicker, simpler and (potentially) more accurate diagnostic tool for cardiac risk demonstrates this fact clearly. If we want to go back to this whole jobs and AI discussion again, we can probably start counting GPs in the number of those whose vocation is likely to be shot down in the automation apocalypse.
Well, yeah, okay, so moths’ brains are pretty interesting. Basically, though, what this comes down to, in my mind, is how you define ‘smartness’. Sure, a moth’s olfactory system is sophisticated and thus rather fascinating. But I dare scientists to see how well a moth does at a game of Go, or at predicting movements in the stock market. That’s not to say there isn’t plenty more to be learnt from studying the brains of biological organisms that can be applied to AI.
5. Artificial Intelligence’s Paperclip Maximiser Metaphor Can Explain Humanity’s Imminent Doom (Quartz)
The paperclip metaphor is a nice way of terrifying yourself about how bad AI could get without even meaning to. Even setting an algorithm a perfectly innocent task could quickly bring about armageddon if human instructions don’t take into account all the ways the algorithm could interpret its task. Like the lesson of Hal 3000, we can’t always rely on an AI to contextualise its tasks from a human perspective.
Pick of the Week
That’s all for now. Until next week…