Why are tech bosses warning about their AIs?
OpenAI CEO Sam Altman and Dario Amodei, CEO of Anthropic, have once again warned the UN Security Council in New York of the dangers of unregulated development of artificial intelligence. They called for internationally applicable control mechanisms to mitigate the potential risks posed by AI. The European press is concerned.
Not too late yet
Der Spiegel is not impressed by what the two AI entrepreneurs, Altmann and Amodei, said before the UN Security Council:
“It's as if soot-covered arsonists were stumbling out of a burning house and explaining to a fire brigade that had reluctantly turned up which fire safety measures now need to be observed and where in the house there would be space for a fire extinguisher. No citizen can feel safe under such circumstances. ... [Politicians] are at least partly to blame for the AI mess and have set too few guidelines for the industry so far. This gap must now be closed as quickly as possible. Unfortunately, there is as yet no sign of a determined political fire brigade willing to put out the AI blaze. But it's not too late yet.”
Nuclear inferno not to be ruled out
Tages-Anzeiger paints an apocalyptic scenario:
“Imagine an AI agent going rogue and hacking into the launch codes for a superpower's nuclear weapons. Dozens of missiles armed with warheads capable of killing millions within minutes are launched. ... It's already too late to deploy interceptor missiles, so a counterstrike is triggered, either by the government or by an AI programmed by the government for precisely such an eventuality. ... Unfortunately, this scenario of mutual automated destruction is not entirely unrealistic: ever since programmes from OpenAI went rogue and infiltrated a company, the possibility of a nuclear war triggered by artificial intelligence can no longer be ruled out.”
Last call for the chicken coop
Perceptions of AI have changed fundamentally as a result of this summer's events, observes data scientist and Green Party politician Svajūnas Plungė in IQ:
“Humankind has never before encountered anything on a par with its own intellect – let alone an intelligence that surpasses the combined intelligence of the entire human race a thousandfold. What's more: we've never before lived in a world where this superintelligence could be given (or secure) all the keys to our critical systems: banks, electricity, water, food logistics, medicine, weapons and warfare. Nowadays everything is controlled by digital systems. So, as ironic as it may sound – we are currently creating a being with god-like abilities and, without realising it, are already reserving a place for ourselves in the chicken coop.”
The machines are not the problem
Tech journalist Sylwia Czubkowska writes in Tygodnik Powszechny:
“All of this has been and continues to be created by people who design, programme and sell this technology themselves. Not by some kind of autonomous AI. It is the people and corporations who are trying to secure money and power from it that should worry us, not so much the technology itself. If we want to stop this predatory form of AI, we must start by understanding the mechanisms behind the concentration of technological power.”