Shivani Ranchod | February 2023
I have an abiding respect for the power of collective wisdom. It’s one of the reasons I love the Time To Think methodologies that we use at Percept – I’m frequently awed by what emerges when many minds come together. ChatGPT poses the question of what happens when we take the possibility of many minds out to its extreme and remove the constraints of time and geography – and more importantly what the relationship is between many minds and collective wisdom.
The Time To Think approach is predicated on inclusivity, inviting everyone in the room into the conversation. The question of inclusivity is particularly interesting for ChatGPT: should all voices and forms of information be equally weighted (and if not, how is this decided)? How do we adjust for the dominance of white, colonial and patriarchal voices? What happens to indigenous knowledge and emergent voices? Not to mention the absence of information that isn’t codifiable data. The assumptions embedded in what constitutes expert knowledge must be interrogated so that we don’t perpetuate systems of oppression nor create new tyrannies. Inclusivity via a large training set does not necessarily equate to neutrality, and it would be dangerous to imagine that it does.
I’m reading the brilliant “Empire of Pain” by Patrick Radden Keefe which is the story of the systematic manipulation of information that drove the opioid crisis. It is a real-life case study of how we discern (or fail to discern) the credibility of information, and how information can be deliberately manipulated to devastating consequence. Perhaps a machine will be more discerning than the doctors hoodwinked by the Sackler family but the truth is that we don’t understand what currently drives ChatGPT’s discernment, and how that discernment may be shaped over time as the technology is monetised and itself manipulated. I wonder if it is the machines we should be scared of, or rather what it is that humans might do with the machines particularly given the threat ChatGPT poses to the wealthy, and to the current global economic order, as it disrupts work as we know it.
We cannot possibly fathom how our wide-ranging collective engagement (and ability to ask intelligent questions) will shape ChatGPT, and the world around us: the future is unknowable and outside of the control of our individual minds. In other words, it is as it has always been.