4 ms·

What about AI soft skills? The ability to explain your model to non-technical people? To picth your solution to executives? To understand business concepts crit

What about AI soft skills? The ability to explain your model to non-technical people? To picth your solution to executives? To understand business concepts critical to increase the value of the product? Dealing with co-researchers from a different field who might have different criterias about what is a valid factor? Dealing with a IT director who thinks he gets AI because he did stats101 over 10 years ago? The value of AI hard skills do not exist in a vacuum. It might be true that doing AI is easier when all the infrastructure and management buy-in is there. Unfortunatly, these are hard to come by, and easy to loose...

tootie 6y ago
Yes. This is a huge gap right now. AI/ML have advanced to the point that complex tasks can be done fairly cheaply. Implementations are easier to find than valid use cases. I've seen more than a few enterprises build giant data lakes and feed them into sophisticated analysis tools, yet the data never seems to inform any decisions. I've seen computer vision used for some cute gimmicks and it's very accurate, but no one could figure out what it was actually good for in a practical application. Same for AR, blockchain and some others. The tech is amazing, but no one knows what to do with it. We need some user experience visionaries more than we need more engineers right now.