A Brief History of AI
For many people, AI magically became ubiquitous overnight. That’s not true, of course, but its adoption by (or infiltration into, depending on your point of view) many industries has been relatively rapid. I want to help demonstrate how it’s changed some industries and explain why quality data is an absolute necessity, even as AI is bringing change with unprecedented speed.
First, some background: the genesis of AI and machine learning can be traced back to the 1940s and ’50s when those terms first started appearing in academic papers. Development proceeded quietly for decades, with a few benchmarks that made news, primarily in gaming:?
Like many technologies, the capabilities of AI are improving at an ever-increasing rate. AI’s abilities have taken off in the past decade or so, surpassing humans in tests of skills like image, speech, handwriting recognition, reading comprehension, and language comprehension. AI applications with commercial potential are improving so quickly that technology just a year or two old may be obsolete: think of image generation or chat.?
领英推荐
But what does it mean when the rubber hits the road? How is this changing the way we do business? Much like the water analogy for data, the effect AI has is very use- or vertical-dependent: we’re all drinking from different spigots.
Even with all AI's advances - automation, insights, efficiency, and others - there’s still no substitute for high-quality data. Computers may be the machines that run AI, but data is the fuel.?
Feed “bad” data into your AI, and you’ll get suboptimal results. Even worse, a generation later, those results may feed back into your AI and further distort the picture - and you might not even know it. This sticky problem, generative inbreeding, is only now appearing on the radar.
In my next post, I’ll examine some AI use cases and how it’s changing the game in various industries, often in very different ways.
Sources: Cited in article