First things first

If we experience a second "AI winter," -- or something even worse, on a geopolitical scale -- please don't blame artificial intelligence researchers. Instead, blame the media and government officials and everybody else rushing to make AI the second coming of the printing press without really understanding the technology or the market for it. AI seems to me an unlikely catalyst for spurring innovation at a national level, and therefore a remarkably bad reason to start trade wars.

I covered some of this in Thursday's newsletter, but I'm inspired to dive a little deeper because I read these three things this week:

The narrative arc is pretty simple to follow: China and Chinese companies are investing heavily in AI, so the U.S. should try to blunt that. Not only does that seem unlikely to actually work -- I'm not sure how you put math back in the bottle, or stop a country from just developing its own microchips -- but it seems like a lot of wasted energy.

I think most people, wherever they live, would rather see their governments invest in making life better and streamlining government programs. And I as I noted last week, there's no particular evidence that AI has to be the answer to any of these things. There are plenty of innovations and research, in fields range from IT to sleep science, that governments could tap into to actually remake their processes and even societies.

Looking to AI as the thing that will revolutionize a nation's economy (or potentially wreck it should someone else have an advantage), is like a person waiting on a miracle pill to help him lose weight and improve his mental health. There are known solutions out there, but making them work is going to require some effort. And stopping the Japanese from eating so much fresh fish isn't going to help you.

And furthermore, where AI is the answer and will be the answer is in areas where there's a problem, or opportunity, for which it just happens to be the best solution. In other words, all the research in the world isn't very valuable if you don't have a killer application and companies or institutions able to harness it. This is why I question all the talk about governments investing billions of dollars to become AI hubs, as if that alone is the answer to their economic futures. In this regard, AI is no different than other technologies.

Maybe you recall all the tech-news stories a few years ago about how Zappos founder Tony Hseih was going to build a second Silicon Valley in Las Vegas. Spoiler alert: It didn't happen and any remnants of the hundreds of millions invested into accomplishing this goal are all but gone. Why? Because it takes more than a will and some money to make something like that happen. It also takes a solid university system, a pool of talent, savvy investors and entrepreneurs, and innovations that people actually want.

So when I see countries like Canada or the UK investing billions of dollars into fostering AI, I'm hopeful for their success but also skeptical about it. (For more on Canada, specifically, listen to my June podcast interview with AI expert Geordie Rose.) Because having AI talent is only a small piece of the puzzle, behind all the things I just mentioned above. I wouldn't bet the farm on AI as much as I'd bet the farm on building the infrastructure that fosters innovation and entrepreneurship across the board.

AI, broadly defined, might very well be the solution to a lot of problems, but that knowledge isn't very helpful to a national economy if it never gets out of the lab or if it's commercialized somewhere else. As the latest step in the evolution of big data, I think AI can be remarkably useful and indeed already has proven itself so. But, man, I wish we'd stop coveting it as the one ring to rule them all, and start thinking about it as yet another tool in a greater effort to just get stuff done and make life better.

Read and share this issue online here.

AI and machine learning






Sponsor: MongoDB

Cloud and infrastructure






Sponsor: Replicated

Data and analytics