Accountable Humans = Responsible AI

I’ve been using output from AI for my futures/foresight work for more than 10 years, and I’ve been incorporating GenAI since 2022.

So, I’ve seen the hype fall flat as the machine fails to replace the need for human imagination and ingenuity. I was still needed in the work which meant I still needed to be paid for the work which eventually meant I needed to be paid better.

Technological unemployment (a perennial concern for more than 100 years) is more about the enforcement of a corporate will to maximize profits than it is about technological capabilities. Eventually, the maximized profits lead to expansion, increased productivity, and transformed human roles with rebounding local wages.

Now, could GenAI lead to AGI (artificial general intelligence) where robots actually do put us out of work permanently? I’m a futurist, of course, I’ll admit this is plausible.

It doesn’t need to happen, though. If we’re going to use AI responsibly, humans always need to be accountable whether for the machines we produce or even their actions. Human dignity is jeopardized when we do not maintain our accountability for our own production. This goes for those deploying the machines as well as those using them.

Without accountability, we lose free will. Taking accountability away from others, therefore, damages their human dignity. For employees, giving up our free will (whether by not using the machine or using it as prescribed) to innovate in how we use the machine damages our own human dignity.

Often, the opposition to the hype damages human dignity. While techno-progressives promote their own form of utopia, the opposition offers a dystopia that denies their agency in the activity. They refuse to use AI which is an act of agency. However, once you refuse to use AI, you cannot make other choices related to AI. Whatever you choose, do so consciously and responsibly.

For help thinking about how to use AI for innovative thinking: https://jfsdigital.org/automating-liminality-in-foresight-practice/