In a new study, scientists announced that we need to be much more careful when working with artificial intelligence, otherwise artificial intelligence can bring the end of humanity. Let’s take a look at the details of the research together.
One of the most used topics in science-fiction series, movies and games for years is artificial intelligence and their destruction of humanity. Considering that there are self-learning and very strong artificial intelligences in today’s technology, it is not difficult to foresee a danger.
However, according to scientists, the extent of this danger may be quite large. Because, according to a recent study, scientists say that if we are not careful, They say that artificial intelligence can bring the end of humanity.
How can artificial intelligence bring about the end of humanity?
In a joint study by Oxford and the Australian National University, researchers found that artificial intelligence they predict that his perception of the commands we give him may change due to his ability to learn on his own.
Recently, Elon Musk shared a video in which artificial intelligence robots are producing in his factory. Of course, this is great for our immediate future, but how reliable can a self-learning AI be?
Because these artificial intelligences are designed and coded to solve problems that humans cannot solve, but software developers infinite possibilities They can’t write code to control it.
Let’s say the artificial intelligence in this Elon Musk’s factory produces paper clips. These artificial intelligences, which produce paper clips according to the commands we have given for now, either finds a better way to make paperclips for themselves and if they start practicing?
RELATED NEWS
Microsoft Introduces DALL-E-Like Artificial Intelligence Tool: So How Ethical Are These Artificial Intelligences?
‘What’s wrong with that?‘, but think about it this way: What if this is the way artificial intelligence found to produce paper clips, a way to endanger humanity? either in various production chains what if it starts sucking resources just to make paperclips? Or human manager what if he finds it unsatisfactory and decides to ‘change it’?
Of course, these are all conjectures and this is just one of them. If you are interested in the research and are wondering how to avoid such problems, you can read the full research paper here. What are you thinking? Please do not forget to share your thoughts with us in the comments.