AI. Uh-oh. The looming nemesis of our species? A discovery comparable to harnessing fire, with immense potential to improve our lives? Or just another tech toy – fun, educational, but not world-changing?
How are we, regional business owners and leaders, to respond to the most durable headline of the past 12 months? Let’s start by recalling a few important facts.
- All technologies are double-edged. We’ve used fire to feed, defend, nurture and comfort ourselves. We’ve also used it for horrifying acts of war and cruelty. The difference lies not in the machine or the stars, but in our choices.
- Close up, what generative AI does is very simple. ChatGPT and its ilk essentially take a series of words, and predict the next one. What gives these machines emergent properties (like seeming to know us and converse with us) is the phenomenal number of predictive elements in their internal networks, and the literally super-human volume of sample data they have digested.
- We can bet with confidence that the hype, on both the promise and the threat, is overblown. The very term “artificial intelligence” is industry hype.
- We are constantly teaching these machines to give us more of what we like. They only “learn” (more neutrally, adjust values in their internal equations) when we reject or correct an output. OpenAI are perfectly clear that when we use their systems, we are helping their R&D.
- They do not form intentions. All an AI system ever does is produce an output in response to an input, which is all that computer systems have ever done, from the beginning. Input – process – output. For a system to destroy people, or truths, or objects, a human must have told it to.
Here’s what I think the potential is: they can support our work in ways we are just beginning to see. Although the architecture is called an “artificial neural network”, and its human inventors were inspired by the actual neural networks inside our skulls, we know that they don’t do what they do, in the same ways that we do what we do.
As with the industrial robots that are already all around us, we have the opportunity to take the special strengths of humans, plus this new tool, and make a whole that is greater than the sum of the parts. Collaborate, not compete – how about that?
Here’s what I think the threat is: they are growing exponentially in complexity and data volumes. They will get better and better at giving us the outputs we accept. We will be less and less aware of when we are allowing ourselves to be deluded. Outputs that seem to us helpful and true will in some cases be clever artifacts, not the real deal.
Here’s my suggestion: let’s learn how to use these new tools to enhance human work and reduce mental drudgery. Let’s stay cautious and level-headed, and be on our guard against those who will use them to do harm. And let’s keep checking in with reality.
What do you think? Are you using AI in your business? If so how? Do you see the threats looming larger than the potential, or vice versa? I’d love to hear from you.
ChatGPT itself, by the way, suggested I describe it as an “elusive magician, capable of both awe-inspiring feats and enigmatic mischief.” I think it kind of liked the tech toy idea 😊.