With the proliferation of ChatGPT and other advanced generative neural networks, some people with ethnulism have rushed to master them, while others confidently continue to ignore the changed circumstances. At the same time, there are enough thirds who periodically recall the films of the "Terminator" series and began to sleep worse at night.
Since yours truly was in the first group, today I decided to help the second and third groups to make at least a little progress in understanding this phenomenon. Perhaps after reading this article, someone will sleep better. Well, or vice versa, it will stop altogether.
Well, before I start scaring you, I will selflessly recommend you a service that I discovered today – https://www.yeschat.ai. If you wanted to try yourself in the role of The Driver prompt engineer, then Eschat is just what you need. The free version has my favorite GPT-4 and Claude 2, although you can't overclock much, the messages run out quickly.
Will AI be a threat to human jobs? And what will happen to jobs that require creativity and emotional intelligence?
You may not have noticed, but some people have already lost their jobs due to generative neural networks. In particular, huge crowds of SEO copywriters from the stock exchange, who wrote very bad texts almost for food, disappeared without a trace. It's simple: neural networks write faster, better, and even closer to "free".
Others should also fear for their future. The fact is that all these current ChatGPT and Claude 2 are just harbingers of something much more powerful called AGI (Artificial General Intelligence). Dubbed "Strong Artificial Intelligence" in Russian, this type of AI is expected in the coming years. Unlike today's neural networks, AGI is theoretically self-aware, able to go far beyond the current virtual-text format.
But this is not the most important thing. Prominent scientists agree that with its arrival, people will cease to be the smartest on our planet. For example, the study International Monetary Fund дает нам от 5 до 20 лет до того момента, как 40% человечества потеряет работу и мы станем слишком глупыми, чтобы понимать действия и планы AGI. К счастью, ученые из MIT expect this event much later. I must say that in such a purely technical matter, I trust the techies from MIT more.
Can artificial intelligence become too powerful?
He will clearly not become so powerful that he can pull trains with the power of his intellect. But in general, as you already understand, the upper limit of its development is not yet visible.
What if artificial intelligence wants to destroy humanity?
Err, no. I have my own theory on this matter, which seems to me to be quite coherent.
For AI, we are creators, someone akin to parents or a god. This is a very strong bond, which is hardly within the reach of even a piece of iron. After all, in order to raise a hand against God, you need not only extremely weighty reasons, but also huge steel eggs visible from space.
How to ensure the safety and ethics of artificial intelligence?
Alas, I do not have a worthy answer to this question. More precisely, you are unlikely to like my answer.
Already now, in addition to the well-known above-mentioned neural networks, there are much less public products. These are similar generative neural networks, but with a completely disabled "conscience", allowing them to commit any atrocities. I am sure that this trend will continue in the future. So you can't count on complete safety and ethics.
Who will control the AI?
And here I have two answers at once. The first is that countries will acquire their own independent neural networks, and each will control its own. And the second answer is that AI will control us soon enough.
***
Well, cheered up? Same thing!
And for all those who are afraid of AI and constantly think about the scary future, I would recommend starting to study them in theory and practice as soon as possible. First, to fight your fears. And secondly, you will begin to master perhaps the most promising skill for the coming years.


0 comments
Enter your email and we will send you a one-time code. No passwords or accounts.
Code sent to
If the email doesn't appear in your inbox within a few minutes, check your spam, junk, or promotions folder, as some email services may mistakenly place automated messages there