Continuing my series of posts about AI and machine learning, a big concern is AI taking people’s jobs. In fact, that’s how I first started posting about it, because AI was being touted as a replacement for writing and creating music and the many fun things in life… rather than AI helping do the dishes or laundry. But despite AI being cited as a reason for jobs getting cut, I have been coming across some takes about AI that say, “Not so fast: even if you wanted to take all the jobs, you simply won’t.”
For example, Adam Ozimek writes for The Atlantic about the history of the player piano –which is interesting in and of itself– but he relates to how it compares to our current time and the automation touchstone that is AI. One of the aspects that Ozimek touches on is how there is a “demand for the human touch.” Besides the medical jobs that require a human touch brought up in last week’s post, what I found interesting is that this is not necessarily a “luxury” requirement. In other words, some experiences demand a human touch for that extra experience, but it made me consider what experiences lose value dramatically without any human touch. Not all automation is equal.
And speaking of not all automation being equal, I know one of the aspects of AI, both as a promise and as a threat, is that it need not simply be the automation or mechanization we’ve seen before. AI can make decisions. It can make judgements. There’s two reasons I’m not dwelling on that here. First, depending on how wonky you want to get, there’s plenty of existing automated processes that, if you pull them apart, you’ll see that some part of the automation used to be a judgement from a very knowledgeable or experienced human — even if the knowledge and experience was simply, “that cord is 13 inches long. I will cut here.” Second, there’s significant pushback on AI being used for complex judgement in today’s processes. Yes, I’d like there to be more pushback, but the backlash is such that I don’t think we can say it’s a given AI will be rendering all your legal and medical decisions tomorrow. (I know this makes some technophiles sad).
But what if AI has supplied the seeds of its own disruption? What if AI can become ubiquitous, but not in the way Silicon Valley envisions in the form of data centers that are already getting some pushback.
Musician and YouTuber Rick Beato talked a few months ago about a technology upgrade in his industry, the music industry, and mused how something similar might occur with AI, especially when it comes to those data centers.
Now, I’ve already had some folks talk about how the commercial AI models, the ones many companies are currently draining their IT budgets on, are far better than the basic public models. And all my exposure to the commercial models indicates that’s true.
But then there’s the draining IT budgets.
I believe the future is still up for grabs — and maybe not by an oddly six-fingered hand.