Ethics and AI

Before we set you off to read today’s article, we need to congratulate Upriver Press author Phoebe Jordan. Her book You, Me, and Autism: How to Befriend, Support, and Work with Autistic People just won a silver medal from the Independent Book Publishers Association in Portland over the weekend. You can read more about the book here. Great work, Phoebe, for writing such a valuable and helpful book!

Now, on to some thoughts about AI and ethics.

Questions about the ethics of AI and publishing hung like a storm cloud over the recent conference for four hundred independent book publishers (mentioned above). This topic also relates directly to the broader question of how AI will reshape our engagement with all types of media, not just books. And that issue is addressed thoroughly by our book Overcoming Information Chaos.

In this Upriver Current article, we show how a humorous animated film provokes (hopefully) serious reflection about modern ambitions to automate everything with AI. The 2024 feature-length animated film is titled Wallace and Gromit: Vengeance Most Fowl, by British writer and director Nick Park.

If you have ever seen one of Park’s works, you know that the lead character, Wallace, is a good-hearted but dweeby inventor whose ambition is to make everything easier, faster, cheaper, and more efficient. His wild contraptions automate every daily task: getting out of bed, bathing, dressing, eating. A jelly catapult is synchronized with his toaster so that he doesn’t have to do the tedious work of spreading jam on bread. Why do anything if a machine can do it for you? That is Wallace’s philosophy.

In Vengeance Most Fowl, which was nominated for an Academy Award, Wallace invents an automated gnome named Norbot. It looks like a kitsch lawn ornament, but it runs on an algorithm designed to rapidly tackle yard and garden work. Norbot’s algorithm is based a few genteel commands like, “neat and tidy” and “no job is too small.”

One day, Wallace takes Norbot out for a test. He goes to a flourishing, beautiful, well-tended garden. He sets Norbot loose, like an AI agent, to see what happens. The bot, in a matter of minutes, trims all hedges, reshapes trees, mows the lawn, and blows all the debris away. “Da-na! Neat and tidy!” says Norbot upon completion. Wallace is thrilled by the time and labor savings.

However, because Norbot lacks any sense of aesthetics or artistry, the garden ends up looking as expected: artificial. Trees look like they were molded in ice cube trays. Spiral hedges look like they were 3D-printed. The grass, not one blade longer than another, looks fake.

Soon, Wallace’s neighbors learn what Norbot can do. Thinking exclusively about how much time they can save, the neighbors start calling Wallace to inquire about hiring his gnome. Wallace realizes he has a business opportunity. He could become rich! He could dominate the entire yard and garden industry. The new business will be called “Gnome Improvements.”

Gromit, who is Wallace’s faithful and highly intelligent dog, is the first to suffer from Norbot’s coup. As the gnome takes over all Gromit’s household roles, the poor dog is demoted.

Unlike Wallace, Gromit is philosophical. He considers the big picture. He thinks beyond technological efficiency. He foresees all the bad stuff that Wallace’s inventions could cause.

Bad stuff, a lot of it, does in fact happen in the film. A criminal character, Feathers McGraw, manages to hack into Norbot’s central operating system. He reprograms the gnome’s algorithm, shifting the system’s ethics to “evil.” Then he commands Norbot to build a small army of gnomes just like him. This is a tech version of rotten indoctrination. This is scaled up hyper-efficiency gone morally awry.

Wallace at first doesn’t realize that the Norbots, about fifty of them, now operate with a heinous ethical framework. He’s just excited that he can expand his business to meet the rising demand for garden services. He believes he can control—and get paid for—the work of a whole company of gnomes while sitting in a comfy chair. Who needs to pay employees when you have an army of robots?

Wallace isn’t the only one who thinks exclusively about saving time by hiring Norbots. His neighbors, the consumers in the story, never think about ethics or downstream negative impacts. They only think about finding ways to do everything cheaper, easier, and faster.

Whatever efficiencies Wallace and the neighbors hoped to achieve are soon awash in chaos and mayhem. I encourage you to watch the film to see what happens.  

Here’s the film’s point about AI and ethics. Park, in subtle and not-so-subtle ways, reminds us that AI systems will be programmed according to the ethics of those who design and operate them. The business plans of AI companies will directly reflect the philosophy and values of the CEO, board of directors, and shareholders. Technology, including social media and artificial intelligence, is never ethically neutral. (A great book that makes this same point is Technology Is Not Neutral by Stephane Hare. See my book review here.)

So, which system of ethics is embedded in the algorithms at Meta, Google, OpenAI, X, Anthropic, etc.? Utilitarian, Epicurean, Hobbesian, Lockean, Nietzschean? Ayn Rand? How are these decisions being made and why? Do the CEO’s even think seriously about ethics, or is everything designed solely to attain maximum profit? I’m not sure, but the answer to those questions, as Park shows us, will determine how AI impacts our lives.

For mere citizens like us, it is probably too late to escape the AI tidal wave. We will have to learn to swim in the riptides. But we also don’t have to be like Wallace’s neighbors whose only motivation is to do whatever is cheap, easy, and fast. We can think about ethics. We can make choices about whether and how to use AI. And we can take a stand against the unethical actions of AI businesses.

We can all be a little more like Gromit.

Quote to Consider

“AI is less regulated in America than sandwiches. You can’t open a sandwich shop without having your kitchen inspected. But you can release an AI girlfriend for eleven-year-olds and that’s fine.” — Max Tegmark, MIT physics professor