Artificial Wisdom
This is a ONE THING article, which is featured in FDI’s monthly newsletter. Sign up to receive this in your inbox.
Author - FDI Founder, Paul Smith
ONE THING is for the busy (speak of the devil) Future Director, in which FDI Founder Paul Smith picks just one thing Future Directors oughta know or do or stew on for the next month. We hope you get something (at least one thing) out of it.
This month’s one thing sees us jump on the AI bandwagon that's been preoccupying most of the internet this past month – ChatGPT. (Hell, it was even an answer in the New York Times mini-crossword the other day).
In a nutshell, ChatGPT is a free-to-use chatbot developed by OpenAI, a US-based startup. You type in a prompt, and it spits out an original response written in conversational English. It’s doing this by pulling on a large language model based on GPT-3.5, simply by predicting the next word in a series of words.
GPT-3 has been around since 2020 (3.5 came out in Dec). But now, content creators on the internet have discovered it. They’ve made it autocomplete famous novels. Draw New Yorker cartoons. Offer stock market advice. The list goes on.
Creators are raising interesting questions about plagiarism (who ‘owns’ a piece of AI-generated text?) and copyright (more an issue for ChatGPT’s cousin, DeepAI, that draws on copyrighted images and artworks to generate new ones - much like it pulled on Wes Anderson movies and Star Wars to make the featured image of this article). What if you asked an AI to come up with a patent? Are we about to see the explosion of a new legal field – Artificial Intellectual Property?
There’s also the chortling over its lack of creativity and authenticity. Clearly, it’s not up to scratch, but as the tech gets more sophisticated, who knows what bridges GPT-4 or GPT-5 may cross.
I’m more interested in ChatGPT’s (and more broadly, AI’s) implications for boards, strategy, risk and their role as organisational stewards. What things should it be making board directors think and talk about? What risks and opportunities does it pose for your organisation?
I see three.
1. Will AI make your human workforce more expendable? The crosshairs seem to be on Customer Service teams right now, generic retail responses being something ChatGPT is eerily good at. But it mightn’t be long before the output of other knowledge workers is as easily emulated. The cost savings are alluring. But my counter to this would be – as orgs replace people with robots, a human-to-human interaction becomes a point of differentiation. It creates goodwill. It says “We value people.” That’s powerful. Another question is, who cops it when a bot offers dud medical, financial, or otherwise legally-binding advice? Culpability should be a big part of any AI strategy a board considers.
2. Should AI be a voice in the boardroom? ChatGPT pulls on many examples of language to accurately predict what the next word in a series of words should be. Which means it’s very good at giving the most predictable response to any sort of question. This could be really helpful in the boardroom. Imagine your board has a difficult decision to make. You ask your AI what it thinks. It gives you the most expected, average, generic course of action you could take. That then gives your board a benchmark, a leaping-off-point, against which to assess the decision you ultimately come to. Now imagine feeding company data into it so that that informs its responses too. Interesting, right?
3. Will AI accelerate your organisation’s need for digital transformation? We already know most board directors lack tech creds. But does the development of AIs like ChatGPT create a legitimate urgency, or at least a convenient excuse, to trigger a bigger boardroom conversation about the need for technological change and innovation within your organisation? It might also be a good impetus to discuss shifting your recruitment committee's criteria toward more tech-savvy candidates.
I’ll leave those questions with you, not ChatGPT.
Image source: Scified.com