Will AI destroy us all? What was discussed during first international summit on safe use of AI?

Around 100 distinguished experts and high-profile guests are attending the first international summit dedicated to the safe use of Artificial Intelligence (AI), which takes place on November 1-2 at Bletchley Park in Buckinghamshire, near London. The purpose of the event is to understand what risks AI brings and how these risks can be managed.

Among the summit participants are US Vice President Kamala Harris, European Commission President Ursula von der Leyen, UN Secretary General Antonio Guterres, owner of SpaceX, Tesla and social network X (formerly Twitter) Elon Musk, Microsoft President Brad Smith, Google DeepMind CEO Demis Hassabis, as well as Sam Altman, CEO of Open AI, the company behind the popular ChatGPT chatbot.

The venue for the summit was not chosen by chance: during the Second World War, Bletchley Park housed a decryption center, and it was here that Alan Turing’s team managed to uncover the secret of the Enigma code used by German encryption machines.

Could AI create biological weapons and destroy us all?

Artificial intelligence could create biological weapons that could endanger the lives of millions of people. This opinion was expressed by US Vice President Kamala Harris during the summit.

“AI has the potential to bring not only good, but also enormous harm: from AI-powered cyberattacks on a scale larger than anything we have seen before, to AI-created bioweapons that could endanger the lives of millions of people. These threats are often called existential AI threats because they could threaten the very existence of humanity,” she said.

According to her, these threats are large-scale and require “action at a global level” aimed at analyzing and eliminating all risks from AI – both for humanity as a whole, and for individuals, and for vulnerable groups of the population.

The US seems to be genuinely concerned about the possible risks associated with AI. President Joe Biden recently signed an executive order requiring developers of serious AI models to share safety testing results with American authorities.

The development of AI should not outpace its regulation

AI technologies should not develop faster than the rules for regulating them are developed. This opinion was expressed by Indian Minister of State for Electronics and Information Technology Rajeev Chandrasekhar during the summit.

“We have learned over the last 10-15 years that by allowing technology to outpace its regulation, we expose ourselves to the toxicity and misinformation that spreads through social media on the internet. We believe that digital platforms must be accountable to the law and ensure the safety of their users,” he said.

According to him, AI opens up many new opportunities for humanity, but such rapidly developing technologies also have a downside.

“We must look at AI through the prism of openness, security, trust and responsibility. We believe that in the future, AI should only be used for the benefit and prosperity of citizens of all countries,” Chandrasekhar said.

Apocalypse according to Musk

A summit on such a topic could not have taken place without the presence of Elon Musk. On the eve of the event, he once again issued a warning that AI could lead to the extinction of humanity. He voiced the same idea during the summit.

According to him, AI poses one of the most serious threats to humanity, a real “existential risk.”

“For the first time, we are faced with a situation where we have something that will be much smarter than the smartest person. It's unclear whether we can control it, but I think we should strive to steer [technology] in a direction that benefits humanity," Musk said.

In the past, Musk has repeatedly spoken about the risks associated with the development of AI.

“There is some probability, better than zero, that AI will kill us all,” he said.

Musk was also among the AI researchers and businesspersons who demanded a halt to the development of neural networks more powerful than GPT-4.

It should be noted, however, that in April Musk registered a company called X.AI, which he said would develop its own AI chatbot called TruthGPT as an alternative to the "biased" ChatGPT.

There are more real risks today

During the summit, many calmer and more restrained statements were made: for example, many experts noted that one should refrain from speculation about unlikely threats in the future and focus on the potential risks that artificial intelligence poses today. For example, AI already poses a very real threat to some professions. In addition, during the learning process he may indeed become biased.

As UK Digital Secretary Michelle Donelan noted, world leaders must work together to “respond to a socio-technical challenge that transcends national borders.” She said AI models “can free people from tedious work and enhance creativity,” but it can also be used for evil purposes—to undermine public trust and so on.

On the first day of the summit, 28 countries adopted the Bletchley Declaration on the safe use of AI.