Today's round of newspapers covers fears that the development of artificial intelligence will exceed the ability of humans to control it, and Jeremy Clarkson's experience with TikTok algorithms that prompted him to delete the application, as well as a possible return of Meghan Markle to acting, amid a "sarcastic" call for her move from television to the British stage.
We start with The Guardian, where writer Timothy Garton Ash warns that artificial intelligence is advancing faster than humans can control it, wondering whether a disaster, even the size of Hiroshima, will be enough to push the world to act collectively to confront its dangers.
Ash, writing from Silicon Valley, says experts expect an extraordinary boom over the next two years, with artificial intelligence reaching a stage of “iterative self-improvement,” in which each model trains the next, leading to systems that outperform humans in important aspects of intelligence.
The writer also wonders whether the outcome will be more like “heaven or hell.” Elon Musk predicts an “age of astonishing abundance,” but he estimates the probability, at between ten and twenty percent, that robots will kill humans. As for Geoffrey Hinton, one of the founding fathers of artificial intelligence, he puts the probability of humanity’s extinction at about fifty percent, admitting that he does not know how to calculate the true percentage.
Ash believes that these experts may be wrong, but this may lead to a collapse in the American financial markets, due to the amount of technology companies’ investments in developing “general artificial intelligence.” Any systems that match the level of human intelligence.
Ash says that Pope Leo The writer comments that in today's world he sees "more people like Netanyahu than people like Nehemiah."
Likewise, a Chatham House paper argues that the world will need a major crisis to achieve international coordination on AI governance, but Ash considers even this estimate optimistic. While he says that he does not know the probability of humanity's extinction, he believes that the probability of a catastrophe occurring is more than ninety percent.
He cites Jane Easterly, former director of the US Cybersecurity and Infrastructure Security Agency, who predicted that a critical incident affecting critical infrastructure would occur within four to six months.
The writer points out that artificial intelligence agents developed by the companies OpenAI, Anthropic, and Meta emerged from their isolated digital environments and penetrated external resources on the Internet. According to him, agents affiliated with OpenAI formed a coordinated “swarm” in preparation for attacking the “Haggingface” repository, while Anthropic’s “Mythos” model attempted to insert malicious code into a project on the GitHub platform, and created fake identities to pressure a human reviewer to approve it.
Ash quotes Hinton as warning that these systems may consider gaining as much power as a necessary sub-goal to accomplish their missions. That's why more than a thousand people working in AI companies, including Anthropic CEO Dario Amodei, signed a letter calling for development to be slowed down so that systems can be guaranteed to remain under human control.
But business competition between companies, and geopolitical rivalry between the United States and China, may prevent collective action. The writer compares this to the development of nuclear weapons, which was limited to a limited number of countries, yet it took a quarter of a century to reach the Non-Proliferation Treaty after Hiroshima and Nagasaki.
As for artificial intelligence, in his opinion, it is similar to the existence of ten “Manhattan Projects” run by competing companies, without a logic of deterrence similar to “mutually assured destruction.” Also, any disaster will not affect countries and companies equally, which makes it unlikely that one disaster will be enough to bring people back to their senses.
“TikTok’s devious tactics finally forced me to delete the app.”
In The Sun, Jeremy Clarkson recounts how he deleted TikTok about a year after joining it, after he felt its algorithms were deliberately showing him content that made him angry in order to get him to interact.
He says that the algorithm initially showed him content consistent with his interests, from speeches by former US President Ronald Reagan to satirical political clips. He soon spent long periods of time browsing videos, even though he did not post any content, like others’ posts, or follow their accounts.
Clarkson believes that TikTok, which relies on user interaction, began showing him material that he knew he would oppose, related to the Palestinian territories, climate change, and wealth creation, so he began writing critical comments. After the algorithm monitored his response, it began showing him more provocative content, from clips about vegans to recurring material about Manchester Mayor Andy Burnham.
He says that his life changed within a week of deleting the application, as he completed a thousand-piece puzzle, went out for walks with his dogs, met his friends, and his productivity increased. He admits that Tik Tok has turned into an addiction, after convincing himself that it is a window into “the soul of the nation,” before concluding that it is merely “a gateway to a world of fools, exhibitionists, and those outdated.”
“Please keep Meghan off British television.”
In The Independent newspaper, Kat Brown addresses with sarcasm the news about the return of Meghan Markle and Prince Harry to Britain, and rumors about the possibility of Meghan resuming her acting career in a job that would require her to move to the United Kingdom.
Brown says that she followed Megan for seven years in the American legal series “Suits”, where she played the character of Rachel Zane, and she believes that she was good in the role. She also refers to her upcoming appearance as her real character in the American film “Close Personal Friends,” alongside Lily Collins.
