Is Artificial Intelligence a Threat to Humanity?

When Geoffrey Hinton quit his job in April this year as a Google top AI scientist also known as the Godfather of AI, he said he was doing so to inform people of the dangers of technology. But don’t we as people have received this warning several times during our lifetime? According to him, he can imagine artificial intelligence technology taking over humanity now. 

But then he is not the only one talking right now about its dangers. As per a survey report given in the link, 36% of AI experts believe it to be true. Do you want to know what more is happening?

An open letter was signed by 28000 people recently and that includes names like Elon Musk, Steve Wozniak, and CEOs of different AI companies alongwith several technologists. The letter written by the Future of Life Institute is asking for a 6-month moratorium which means they want the world to temporarily cease all AI activities for 6 months. It means they do not want any more AI development or to say advanced AI development during this time.

Also Read: What are Smart Rings? Advantages & Disadvantages

What are the stepping stones which are alarmingly destroying the world’s peace on this matter?

1. AI development is proceeding very rapidly and at this speed, people fear it won’t be 5 years before they will be facing a challenge.
2. The coming of large language models (LLM) is posing just one chance to humanity to get everything right. If not we might not be able to survive it.
3. If LLM goes on to improve everything so rapidly, it will soon get to the stage of AGI (artificial general intelligence) which means machines will no longer need human intervention to improve AI.

Also Read: Bill Gates Prediction for the Future of AI

If you know Google’s Alpha Zero model that’s already touching this stage it has learned to play the best Chess ever in just 9 hours from the time it was switched on. It defeated not just other AI models but even the best human Chess players.

GPT-4 already has a spark of the AGI built in it and did better than 90% of human candidates in the Uniform Bar Exam. These tests are reasoning evaluations and therefore Microsoft’s Bubeck and his team feel they are witnessing the earliest version of AGI with GPT-4 system right now.

Also Read: “Is AI the Game Changer of the Future?”

Also known as the control or alignment issue as per some and as Hinton says people to look at what AI was five years back and what it is now. Nobody knows what AI is capable of or is already evolving internally. Regulation is very crucial at this stage as per Hinton who feels if this stage of regulation is missed humanity may not have a second chance.