Artificial intelligence, or AI, seems to have finally arrived. With the introduction of his ChatGPT last November, millions of people suddenly discovered that AI is more than just a research activity. The range and sophistication of ChatGPT’s answers to a wide range of questions is, quite frankly, quite astonishing.
AI already exists in many places you may not be aware of. Google Translate has been using AI for years and while it falls short of human translators, it’s still very good. The technology Pandora uses to customize its music is a form of AI, similar to the technology behind Tesla’s self-driving cars. In my own field, the program AlphaFold2 was a real breakthrough in a scientist’s ability to predict protein structure.
But in addition to these seemingly beneficial developments, a number of concerns arise. As AI gets better and better, can we continue to trust it to make decisions for us? Mistakes can be much more expensive if you’re driving a car.
And there are far more risky tasks that AI could soon perform. What if the military started using AI on the battlefield? It might be tempting to say no, of course we wouldn’t, but what if other countries started using AI in this way? can you just ignore it?
Honestly, at least I’m not worried yet. I have a secret to confess. I got my Ph.D. Papers on AI in a field called machine learning. I haven’t done any AI research in the last 25 years, but I think I understand at least a little of the fundamentals behind the technology.
But thanks to ChatGPT’s attention, there is suddenly a big concern that AI is a danger to society. No one really knows why. There is great potential and great danger in what we are doing.”
Of course, politicians don’t know what AI technology is, but I couldn’t help but laugh when the White House press secretary said: No, he doesn’t know, but I’m skeptical if he CEOs of Google and Microsoft know how it works. But you don’t have to understand technology to understand the dangers.
Just recently, one of AI’s early pioneers, Geoffrey Hinton, spoke out to warn the public that AI has reached a point where “something really bad could happen.” Not surprisingly, other experts disagree, including German scientist Jürgen Schmidhuber.
FTC Chairman Lina Khan says AI can be effectively regulated, and it probably can. But I’m not sure regulation will eliminate the dangers Jeffrey Hinton is worried about.
Much of the media concern seems to be about using AI to spread misinformation. For example, AI can help create “deep fakes”. This can be used to convince people that, for example, an election was stolen even though it wasn’t. This is a legitimate concern, but it’s happening even without AI.
The better question my family asked me is should I worry about AI making decisions on the battlefield? I thought this question should be taken directly to the source, so I asked ChatGPT itself. The answer I got was very clear and direct. I quote the full text here.
“Yes, people should be concerned about AI making decisions on the battlefield. There are several reasons for this.
First, AI lacks human empathy, judgment, and intuition. In complex and dynamic environments like the battlefield, human decision-making is often guided by experience, intuition, and broader contextual understanding. AI lacks the ability to consider these intangible factors and may base decisions solely on pre-programmed algorithms or data inputs, leading to unintended consequences and even catastrophic errors. There is a nature.
Second, the use of AI in military decision-making raises serious ethical concerns. For example, AI systems may be programmed to prioritize the protection of friendly forces at the expense of civilian lives, with potentially devastating consequences for non-combatants caught in firefights. I have. Additionally, AI systems are susceptible to hacking and manipulation, which can lead to unintended consequences and loss of control over the decision-making process.
Finally, there is a risk that reliance on AI systems could lead to a decline in human decision-making capacity, lower the overall quality of military decision-making, and increase the likelihood of error and unintended consequences. ”
So, ChatGPT itself gives some reasons why it’s a bad idea to let AI make decisions that could cost human lives. We need to listen to what it is telling us, because as it becomes (and will become) more intelligent, it may decide to be less honest in the future. I have to say that the reasoning here is well written and persuasive.
One of my concerns is that the term “artificial intelligence” itself is too broad. We have to worry about giving technology the power to control our lives, no matter what someone calls it. We may want to listen.
Follow me please twitter. check out my website.
