Thinking Out LoudLast week, Sam Altman – the CEO of OpenAI, the company behind generative artificial intelligence chatbot ChatGPT – was fired and a few days later, re-hired
The whole thing took around 100 hours. Most of us have open cartons of milk in the fridge that outlasted the story. But the short storm reflected deeper currents. The artificial intelligence community is divided.
The zoomers encourage the unrestrained development and deployment of AI to tackle our biggest challenges: poverty, climate change, productivity and inequality. The gloomers urge a more cautious approach.
They’re concerned that AI puts efficiency ahead of truth and fairness. They’re worried it can be deceptive, that it could even destroy humankind. And they don’t trust humans to navigate its risks. Dehumanisation. Deception. Destruction. Fair concerns.
Especially given that Altman himself says his company has developed AI that’s too dangerous to share publicly. Mary Shelley’s novel, Frankenstein, tells the story of scientist Victor Frankenstein, who builds an artificially intelligent humanoid.
The creature – initially well-intentioned – later turns against his creator, becoming a monster who leaves a trail of suffering. Parallels with modern AI debates are obvious, but they often miss something.
The creature’s dehumanisation, deception and destruction is caused by people, reflecting aspects of the human condition. The monster is synthetically made but his moral brokenness is organically human.
People often make the mistake of assuming that Frankenstein is the name of the monster, not the scientist. Those who have read the novel know that’s not the case.
But in another sense, Frankenstein – the scientist – was the monster after all, the one who couldn’t be trusted with technology. The debate around AI might be about two visions of the future.
But whichever path we tread, we face the inevitability of human fallibility, and nothing artificial can help us with that.