It is hard not to be impressed at how quickly AI technology has developed over the past 10 years. Some applications such as the ability to manipulate or create graphics based on a few simple human-language commands, attract our attention (eg FaceApp on Android is simple and fun). Cars that drive themselves are even more impressive, with the Google spin-off, Waymo, being the market leader and trialling the driverless cars in Phoenix, Arizona and LA, but the technology doesn’t seem to be quite good enough, yet. One worry is that a hacking event or even a software fault such as the one affecting UK Air Traffic Control in 2023 could see our roads jammed with AI cars that simply stop working, or worse, cars that crash.
It is easy to talk about AI talk about it as if it is a form of magic, and this makes it easier for those who intend to make money out of AI such as Elon Musk & Peter Thiel to make optimistic predictions about AI robots working as servants in our houses. People trying to sell you something should be treated with caution.
Artificial Intelligence doesn’t blindly follow programmed human instructions. AI computer systems learn from trial and error, similar to the way that people learn and this has allowed for some rapid advances. Science fiction addicts will remember stories where the computer evolves and becomes so intelligent it decides to take control away from the lesser humans and rule the world. I would suggest that in reality, the problems with AI will come from a human source.
Last week when Geoffrey Hinton won the Nobel Prize for his work on AI, he spoke about the advantages and problems of AI.
“It will be comparable with the Industrial Revolution, but instead of exceeding people in physical strength, it will exceed people in intellectual ability.
“In areas like healthcare it’s going to give us much better healthcare. In almost all industries it’s going to make them more efficient. People are going to be able to do the same amount of work with an AI assistant in much less time. It’ll mean huge improvements in productivity.
The Industrial Revolution Comparison
I think this comparison with the Industrial Revolution is particularly apt. When we created, in the 1800s, machines that had the strength of 100 horses, which could do the work of many men the problem was not that the machines would become all powerful or run amok. The difficulty was that the people who owned the machines became incredibly wealthy and use their power and wealth for selfish purposes, destroying the lives of the skilled workers that machines replaced.
It should be remembered that the Luddites, who opposed mechanised weaving in ‘dark, satanic mills’ and broke up the machinery were stopped when capital punishment was introduced by the British government. (Yes, some people were hung and other transported in the early 19th Century for breaking machines because the machines mattered more to the rich and powerful than the lives of the poor.)
The industrial revolution is remembered as good; it brought us factories, trains, iron clad ships and eventually aeroplanes, but when those inventions were used for imperial conquest and mechanised warfare the results were concerning.
I fear we should expect similar problems with AI because a small number of very rich people will benefit enormously from the use of AI to process information in a way that replaces and controls the current work force. Their aim will be to reduce the costly requirement to have human workers.
Particularly in medicine at the moment, some highly trained people are finding that AI helpers are being developed to improve and ‘speed up’ what they do and possibly replace them. This is seen as beneficial at the moment because in medicine there is always more work, but what about other careers.
Workplace Supervision
I would invite you to look at some of the AI software used to monitor staff and customers in a coffee shop. AI software can track in real time the amount of time each customer spends in the shop, how many cups of coffee each member of staff serves and much more. Some will feel this is no different, or perhaps even better than having a manager always present and watching, but anyone who has worked in customer service will know that constantly being watched by a manager is stressful. In addition, there is no doubt that such software increases considerably the power held by the management of the business and weakens staff independence. Is this a good or a bad thing?
AI is Still Primarily Information Processing
For very good reasons we have laws controlling how much information anyone is allowed to keep or process about us – remember that every one of us will have secrets that we do not want shared. Unfortunately, as the ability of AI increases people will be able to bypass some of those safeguards.
If you look online you will find Harvard students were able to use Meta’s Ray-Ban Smart Glasses to create an AI Surveillance system that could identify people on the street and supply detailed information to the wearer about them. Imagine how useful this would be to a predator seeking vulnerable victims.
AI in Warfare
Already in Palestine we see AI being used in the ‘Is Daddy Home’ system to automatically track targets for assassination, sometime using AI drones to kill the target. Is this a good or bad outcome?
AI Regulation
As was the case during the Industrial Revolution, we will all see some of the advantages of the new AI technology, but those owning the businesses and implementing AI systems will see their power and wealth grow extraordinarily. There are stories from the USA about very rich people like Elon Musk and Peter Thiel trying to influence Donald Trump to make promises not to regulate AI development. You might know that Rishi Sunak paid American company Palantir, co-founded by Peter Thiel to take charge of our NHS data, with all the data mining opportunities this opens up, with little to no regulation on how this data is used. https://www.theguardian.com/society/2023/nov/20/nhs-england-gives-key-role-in-handling-patient-data-to-us-spy-tech-firm-palantir
During the Industrial Revolution we needed regulations, introduced through union pressure, to protect workers and to regulate the power of the industrialists. Similarly, during the AI Revolution, I suggest we will need regulations to protect ourselves from abuse by the people who control AI – the stories about autonomous scary robots taking over is just a distraction. As always, it is other people seeking power and wealth through the abuse of AI that we need to watch.
Arnold is a retired teacher from Belfast.
Discover more from Slugger O'Toole
Subscribe to get the latest posts sent to your email.
