First things first

The European Union and its member countries are waking up to the fact that the United States and China are ahead of them in artificial intelligence research. Efforts to bridge this gap are picking up, including this week with the European Commission prepping an AI strategy and several countries calling for the creation of an AI equivalent to CERN:

However, this might be a whole lot of overkill. At least, the focus on AI might be overkill. It's obviously important to fend off brain drain and to conduct research into how data can save lives or improve citizens' lives. I just don't know that AI is the only, or best, way to do some of these things.

Kind of like how some people get when talking about autonomous cars, governments and even businesses have gotten ahead of themselves worrying so much about having AI strategies. I'm not going too deep into this right now (because I'm traveling; maybe for Sunday's newsletter), but there are a lot of infrastructural pieces that need to be in place to support these efforts over the long term. And, economically speaking, it's debatable just how valuable AI researchers will be compared with entrepreneurs who understand how to turn those advances into products.

More to the point, though, it just seems like governments and businesses obsessed with AI are trying to run before they can walk. Maybe focus some attention on cybersecurity and generally making processes more efficient. Modernize that infrastructure (physical and IT) and actually put a modern data-processing system in place.

Put some real effort into Big Data, Data Science and Data Ethics 101 before jumping into the master courses in AI. Figure out what you even mean by AI, because there's a lot that can be done today (and could have been done over the past decade) that's probably really useful but isn't nearly as futuristic as much AI talk sounds. In my latest podcast, linked to below, ZIllow's chief analytics officer explains how the company got the Zestimate down to a 4 percent error range using various machine learning methods, including computer vision. I'm sure the world's governments and companies have problems they could solve using similar techniques.

AI is going to help solve some problems and cause some others, but the truth is that while we understand what passes as AI today, we don't actually know what a super-intelligent system would look like, much less how it would behave. But even right now we can address a lot of serious issues, and make some serious progress, using existing technologies. If you can't make happier citizens or fatter profits today, AI isn't going to help you tomorrow.

P.S. I am traveling at the moment, so if it seems like I didn't annotate as many of the links below as normal, that's why.

Read and share this issue online here.

Sponsor: MongoDB

The ARCHITECHT Show podcast

AI and machine learning

Sponsor: Replicated

Cloud and infrastructure

Data and analytics