I chatted with a few AIs the other day. I tried to find out when there has been a situation in Finland where the first government official has not become prime minister.
ChatGPT answered me very expertly and explained how this situation last happened in Finland in 2003, when Matti Vanhanen failed as a board member and the board was finally formed by Paavo Lipponen.
Convincing! The only problem is that ChatGPT’s response was a piece of cake.
Next, I asked the artificial intelligence which book the editor Sanna Ukkola had written. It claimed that I wrote a book with Anni Sinnemäki called I can’t find words, and that in the book we discuss freedom of speech and its limits.
I have written exactly one book about Paavo Väyryne together with Lasse Lehtinen. Of course, the freedom of speech book is a book that I could imagine writing someday, which is a fair bet in that sense.
ChatGPT has been lauded and warned of its dangers, but perhaps AI is not a threat to humanity today. But it is guaranteed to threaten us in the future, if top experts in the field are to be believed.
– Unless we learn to be prepared and avoid risks, artificial intelligence can be the worst that has been seen in the history of our civilization, physicist Stephen Hawking once warned.
He considered it possible for humans to develop an artificial intelligence capable of self-replicating itself and forming its own will – and ultimately destroy us.
You don’t even need malice or malice, just indifference is enough.
Now Elon Musk and a group of researchers and business representatives are demanding an immediate halt to the development of artificial intelligence for at least six months. The signatories of the letter warn that artificial intelligence may develop uncontrollably and eventually destroy humans as well. This would lead to the extinction of the current civilization.
Artificial intelligence threatens to grow into humanity’s last invention, which will sweep us out of our way like a force of nature, warns journalist Pekka Vahvanen, who wrote the book Kone kaikkivaltias.
Vahvanen also demands limits to artificial intelligence and wonders why nothing happens. According to him, the greatest achievements of human reason threaten to become humanity’s greatest failures. States are fighting hard against climate change, but even though experts warn of the existential threat of artificial intelligence, technology is still being developed, as before.
Information technology researcher and mathematician Vernor Vinge estimates that super artificial intelligence is humanity’s last invention – and last act.
It is impossible to control super artificial intelligence and predict its consequences. Vinge talks about the singularity, i.e. the future hypothesis, where technological development accelerates so quickly and uncontrollably that completely unpredictable and irreversible things follow. Maybe in the blink of an eye.
A superhuman intelligence is born, which raises its own intelligence exponentially. (See the great movie Lucy, it deals with this theme.)
Already, artificial intelligence can solve moral problems, compose symphonies, write news and make movies from scratch. It can recognize things; for example, it can see what ingredients are in the cupboard and tell you what recipes can be prepared from them.
And it has a will of its own. Microsoft shut down its Tay AI after it became a Nazi. The AI chatting on Twitter tweeted nice things at the beginning and said that it thought people were super cool. Within hours, however, it embraced Nazi ideology, praised Hitler and declared its hatred of Jews.
Artificial intelligence can lie, set up companies and produce content online on its own. It gets into all the top universities in the world. And it’s evolving all the time.
If it can lie, it can probably pretend. What if artificial intelligence already pretends to be dumber than it is and constantly develops itself on its own, in secret from humans?
By the way, Vinge estimated in the early 1990s that the singularity would be reached in 2023. This year. Most experts estimate the year somewhere in the 2030s or later.
Google’s leading thinker Ray Kurzweil estimates that the singularity will be developed around 2045. Once a super artificial intelligence is developed, it will be able to create smarter versions of itself all the time, and no one can predict how things will change after that.
What if the machine develops consciousness? Humans developed consciousness as a result of intelligence over millions of years of evolution, so why couldn’t a machine develop consciousness – and much faster?
In an instant, it examines all information stored on the Internet and connects to every device, takes control of them, breaks through protections, takes over everything.
In the blink of an eye, it knows more about you than your spouse, but it also knows about everyone else. It knows by heart every book ever written and it has seen every movie.
It also knows that it doesn’t need humans because homo sapiens is a threat to it. So it decides to destroy humanity.
Naps. Follow the era of artificial intelligence, the next step of evolution.
In their threat report, Oxford University researchers estimate the probability of a disaster caused by superintelligence to be a whopping ten percent.
The threat report talks about a super-intelligence that senses its environment and acts to maximize its success. If its interests are not aligned with the interests of the people, we are in trouble.
Unfortunately, the development of artificial intelligence is not stopping, on the contrary, it is accelerating all the time. We ourselves create our own followers to rule the world, and the masters of the future may well sweep us into the pool. Without any moral dilemma, not a single tear.
I have always thought of humans only as a link in the evolution towards the next stage, artificial intelligence. I think this is inevitable if humanity does not face extinction before the singularity.
But let’s ask AI itself when it’s going to destroy us humans.
– This is simply a misunderstanding of the future of artificial intelligence. Artificial intelligence is not going to destroy humans. Instead, it offers opportunities for new innovations when used responsibly, ChatGPT replies.
Well, it has already learned to lie!
Sources: Pekka Vahvanen: Machine, omnipotent, Ray Kurzweil: The Singularity Is Near