Have we unleashed a monster? AI has wormed its way into every area of our lives. Now one of its inventors says this is our last chance to stop it turning the world on its head.
'Godfather of AI' quits Google with warning
Have we unleashed a monster? AI has wormed its way into every area of our lives. Now one of its inventors says this is our last chance to stop it turning the world on its head.
Throughout history, many geniuses have regretted their own inventions. Robert Oppenheimer, inventor of the atomic bombA bomb that rapidly releases nuclear energy. , spent his later years campaigning against nuclear weapons. Alfred Nobel created the Nobel PrizeOne of a set of prizes, laid out in the will of Swedish chemist Alfred Nobel, given each year to people who "have conferred the greatest benefit to humankind". to redeem himself for the invention of dynamite. The man who designed the TV, Philo Farnsworth, banned his own son from ever watching it.
Now they are joined by another great scientific mind. Dr Geoffrey Hinton, who is seen as the godfather of AI, has resigned from Google, saying he wishes he could turn back the clock on his life's work.
He warns AI chatbots like ChatGPTAn Artificial Intelligence chatbot released in November 2022. will soon be more intelligent than humans. This is because, he claims, scientists are only now waking up to the fact that artificial intelligence is quite different from human intelligence.
A digital system is made up of thousands of copies of a single model. All of these copies learn separately but share their knowledge among themselves. So a single AI can learn at the rate of around 10,000 people.
Right now, this is not too much of a problem, because while AI knows a lot, its reasoning power currently lags behind that of human beings. But this could all change soon.
Hinton fears that it would only take a single "bad actor", like Vladimir Putin, to give AI the ability to create its own subgoals for this reasoning capacity to accelerate.
The AI might then decide, for example, that in order to achieve its overarching goal, it needs to achieve a subgoal of developing yet more power and intelligence.
Soon, for the first time in history, we could be sharing the planet with a species that is smarter than we are. There is no way of predicting how that would pan out.
And even if AI does not actually turn against us, it can still do a lot of damage. It would hardly be the first time that an apparently innocuous invention has shaken the world to its core.
In around 1440, Johannes GutenbergA German goldsmith, inventor, printer, and publisher. His invention of the printing press in 1440 meant that books could be mass-produced and knowledge could spread around the world. Many historians see this as the beginning of modern history. introduced movable-type printing to Europe. The result was an explosion in the number of texts. People began reading about religion and politics. Some began to criticise their governments and the Church.
By the time anyone had got to grips with the revolution this new technology had brought, western Christianity had split and Europe had been devastated by religious and civil war.
AI systems could have the same effect. They are already threatening to make many workers redundant.
This week, tech giant IBM said it would stop recruiting for around 7,800 jobs, because the roles could be performed by AI instead of humans.
And in the USA, writers for TV and film are threatening to go on strike, in part because they want to ensure studios do not use AI chatbots to write scripts, which would put them out of work.
It is estimated that AI could take over 300 million jobs.1 The people laid off from those jobs are unlikely to take it lying down. Just as the printing press set in motion centuries of warfare and bloodshed, some fear AI could be opening the door to a new age of turmoil and revolution.
<h5 class="wp-block-heading eplus-wrapper" id="question"><strong>Have we unleashed a monster?</strong></h5>
Yes: AI is advancing at a pace beyond our understanding. No part of society or the economy will go untouched by it. The world will never be the same again.
No: AI has its dangers, but we can keep it under control. Hinton praised Google for its "responsible" use of AI. Bad actors might want to make bad uses of it, but none of them currently has the ability to make a sentientAble to perceive or feel things. AI.
Or... AI is not a monster any more than the printing press or the TV. But like those inventions, it will change our world in ways we are currently unable to predict. That inability to know is what makes it truly scary.
Atomic bomb - A bomb that rapidly releases nuclear energy.
Nobel prize - One of a set of prizes, laid out in the will of Swedish chemist Alfred Nobel, given each year to people who "have conferred the greatest benefit to humankind".
ChatGPT - An Artificial Intelligence chatbot released in November 2022.
Johannes Gutenberg - A German goldsmith, inventor, printer, and publisher. His invention of the printing press in 1440 meant that books could be mass-produced and knowledge could spread around the world. Many historians see this as the beginning of modern history.
Sentient - Able to perceive or feel things.
‘Godfather of AI’ quits Google with warning
Glossary
Atomic bomb - A bomb that rapidly releases nuclear energy.
Nobel prize - One of a set of prizes, laid out in the will of Swedish chemist Alfred Nobel, given each year to people who "have conferred the greatest benefit to humankind".
ChatGPT - An Artificial Intelligence chatbot released in November 2022.
Johannes Gutenberg - A German goldsmith, inventor, printer, and publisher. His invention of the printing press in 1440 meant that books could be mass-produced and knowledge could spread around the world. Many historians see this as the beginning of modern history.
Sentient - Able to perceive or feel things.