Glass Half-Full (of Something)
Are we too optimistic about the benefits that software can bring us?
It is true that at The New Stack, we like to elaborate on the possible uses of the technologies we cover and how they could help you — as an administrator, developer, DevOp or architect – get your systems to the ideal state a little bit faster. And though we always look for the (often hidden) tradeoffs of any new technology, we figure that if the technology wasn’t worthy of your interest in the first place then we won’t waste your time telling you about it.
But are we, as an industry, too optimistic? This was the argument that Harvard Professor James Mickens made in a 2018 USENIX conference. In our rush to embrace the supposed benefits of new technology, we forget to turn a critical eye to the latest technologies we hype. And we, and our users, suffer the consequences.
Take machine learning for example. ML is “the oxygen that Silicon Valley is trying to force into our lungs,” Mickens told the audience. ML developers rarely have the mathematical understanding of why an ML system produces the results it does, let alone a solid understanding of the ethical or even legal issues ML brings up. As examples, we just look to recent AI disasters such as Microsoft’s Tay or the ongoing issues around bias in AI systems.
Or, you could look to the recent debacle around the U.S. Iowa democratic caucus. Whatever your political views, I’m sure you can see how un-earned optimism around the tallying app had led to many questions and frustrations — not only about the results but even about the whole caucus process itself.
We’re not against technical optimism, but at the same time, thanks to Mickens, we’ll take a closer look at the technologies that promise a bright future for all.
|