AI: Prometheus or Frankenstein?
Since ChatGPT has been introduced to the public, tons of comments, posts, etc. have been published. The first step was amazement, but it has quickly opened the way to fears, some controlled, some totally uncontrolled. ChatGPT and more generally the progress of Artificial Intelligence (AI), makes us revisit the great myths of our culture.
According to Greek mythology, Prometheus stole fire from the Gods and gave it to humanity in the form of technology, knowledge, and more generally, civilization. In other words, sometimes a great disruption against established habits and values is the best source for human progress. However, the second step is the punishment of Prometheus: Zeus punished him to eternal torment, an eagle was sent to eat his liver, until he was relieved by an intervention of Heracles. Prometheus is regarded as an incarnation of geniuses who develop technology breakthroughs improving humanity.
A more modern myth is the myth of Frankenstein, essentially coming from the XIXth century novel by Mary Shelley, titled "Frankenstein or The Modern Prometheus." The story is supposedly built upon the ancient Golem legend, a Jewish folklore creature made of mud that comes to life. In both cases, the Creature rebels against its creator. The Frankenstein myth is often used as a criticism to unrestrained scientific experimentations. In this context, in the XIXth century understanding,
one would consider mankind was meddling with powers that rightfully belonged only to God; powers that mankind couldn't hope to understand or control; and powers that God would punish them for meddling with.
ChatGPT, now in its GPT-4 version was designed and built by humans using advanced technology. As it uses natural language processing (NLP) algorithms to generate responses to users' inputs, it may resemble to human dialogue even it is not bound to get incarnated soon. The creation of intelligence by humans always brings us back to the Frankenstein myth. Even Sam Altman, the CEO of OpenAI, the company that developed ChatGPT, has warned that the technology comes with real dangers as it reshapes society, according to The Guardian. He adds that he is calling for regulation to guard against potentially negative consequences for humanity. “I’m particularly worried that these models could be used for large-scale disinformation,” Altman added. “Now that they’re getting better at writing computer code, [they] could be used for offensive cyber-attacks.” But despite the dangers, he said, it could also be “the greatest technology humanity has yet developed.”
Calls for regulation of AI have been coming from various stakeholders already for years. Some major players, including Google, BT (British Telecom) and Microsoft, which is integrating ChatGPT with Bing, have announced attempts for self-regulation. Governments in the US, UK, China and in the EU are working on regulating Artificial Intelligence. In the EU, the proposed AI Act, assigns applications of AI to three risk categories.
· First, applications and systems that create an unacceptable risk, such as government-run social scoring of the type used in China, are banned.
· Second, high-risk applications, such as a CV-scanning tool that ranks job applicants, are subject to specific legal requirements.
· Lastly, applications not explicitly banned or listed as high-risk are largely left unregulated.
Given the speed of progress of AI, one may wonder if regulation will ever be sufficient to prevent abuses of the technology by major corporations or authoritarian governments. And regulation can hardly stop the progress of technology: “There’s always a reason given to validate these Frankenstein-like experiments,” said Charles Krauthammer, an American political columnist.
So, is AI closer to Prometheus or to Frankenstein? We cannot leave it to Parliaments to decide, this discussion is too important to be left to politicians, it should be a global cross-society debate.