Getty ImageAs I write, many powerful people all over the world are preparing to head to Paris. And it’s the feeling that many of them are holding their breath.
On Monday, representatives from 80 countries, including world leaders, technical bosses, academics and other experts, will gather to discuss current progress and future, despite the city’s 125-year-old Grand Palais. We will meet at the two-day global summit for the event. A goal for rapidly evolving, highly destructive technology that is artificial intelligence.
It may be on the official AI Action Summit agenda, but you may be smoking a fire at this particular talking shop, Deepseek.
Horse racing fans once said that the night before a big race, everyone was the winner.
And China has dramatically blown away the AI competition, and in a new direction with Deepseek, its super efficient and super violent AI assistant, the position of poles occupied by the US AI sector is enormous and enormous. There is a sudden sense before the summit, which is occupied by wealth. AI infrastructure may not be so obsolete after all.
Professor Gina Neff of Minderoo Technology and Democracy Center at Cambridge University says there is now a “vacuum for global leadership in AI.”
Professor Dame Wendy Hall, a graduate of the University of Southampton, agrees. “Deepseek made everyone realize that China is a force to consider,” says computer scientists.
“There’s no need to go with what big companies on the West Coast are saying. We need global dialogue.”
On that front, the timing of the summit never improved.
Europe is also spying on the opportunity to make a new bid for the AI Crown. One official of French President Emmanuel Macron described the summit for journalists as a “wake-up call” for France and Europe, adding that Bullock “can’t go past it” the AI revolution.
Other countries are also aware of a potential shift in AI power in the air. India’s Prime Minister Narendra Modi confirmed his attendance at the summit.
The US has sent serious firepower as its own defensive signal, including Vice President JD Vance, Openai CEO Sam Altman and Google’s Sundar Pichai.
Elon Musk doesn’t particularly exist on the official guest list, but he definitely has something to say about it all, whether he’s there in person or not.
British Prime Minister Kier Starmer is also reportedly apart.
Getty ImageThere were two previous summits. It was first held in the UK and the second in Korea. Since the first time the AI world came together, and since they gathered on the stairs of Bletchley House’s historic mansion in November 2023, a lot has happened, and they are trying to maximize the benefits of AI while minimizing risk. I promised to do that. For one thing, half of the world’s population has since been involved in polls.
Wu Cha-Ou Yi was present at Bletchley Park, China’s Deputy Minister of Science and Technology. However, there were whispers that he was still at the length of his weapon, citing national security.
In Paris, by comparison, I expect China to be an honorary guest. The country reportedly has sent out Ding Xuexiang, a close ally of President Xi Jinping, one of its most senior leaders. There is also talk of whether Deepseek creator Liang Wenfeng will join him.
I asked ChatGpt to list some of the notable milestones of AI since the Korean Summit in May 2024. Deepseek was not cut.
Of course, there are far more to it than generative AI like both deepseek and chatgpt – tools to create content such as text, images, videos and more. As consumers, it may be the most widely accessible for us. However, there are also AI tools that can detect symptoms of disease, model climate change, and develop new prescriptions for drugs. It all falls in Paris.
Furthermore, the story of David v Goliath, whose Deepseek’s story hangs, deserves even more scrutiny. Dario Amodei, the human boss of AI company, wrote a compelling blog about whether Deepseek was actually built with just a small portion of the cost of its US rivals.
We know it was built on their shoulders: using a number of nvidia chips (presumably old due to US sanctions) and some OpenSource AI architectures developed by Meta I will. Furthermore, Openai complains that its rivals are doing their own work (considering that Openai products “are willing to generate the style of individual human creators, the contact information for the creative industry is here I was fascinated by the irony)).
Nevertheless, Deepseek has managed to shake up the AI sector in ways that even AI itself may not have predicted. And it wiped out a lot of money from the value of the biggest players in the process. It will almost certainly be a big topic of conversation in those Parisian conference rooms.
Getty ImageThere is another theme that runs through the AI Summit, which is worth noting.
The first summit had the word “safety” in the title. Some felt that the event pushed the story so hard that it terrified people with a dark story of existential threats.
But it has not fallen completely off the agenda.
As a subject, AI security is a rather wide church. It includes any risks, including the generation and spread of misinformation, displaying bias and discrimination against individuals or races, the ongoing development of AI-controlled weapons by multiple countries, the possibility that AI could create unstoppable computer viruses. can be related to.
Professor Jeffrey Hinton, often referred to as one of the godfathers of AI, calls these “short-term risks.” They may be debated in Paris, but he argued last week that it is unlikely to win strong international collaborations in the long term on BBC Radio 4’s today’s programme.
The big scenario he really believes everyone will put together is the outlook that AI will become smarter than humans and want to gain control.
“No one wants AI to take over from people,” he says. “The Chinese people actually wanted the Chinese Communist Party to run shows more than AI.”
Professor Hinton compared this unforeseen situation to the height of the Cold War. This was “almost successful” in the United States and Russia working together to prevent a global nuclear war.
“There’s no hope of halting (AI development),” he said. “All we have to do is try to develop safely.”
Professor Max Tegmark, founder of the Future of Life Institute, also shares a harsh warning. “We’re developing amazing AIs that help humans, or we’re developing out-of-control AIs that replace humans,” he says.
“Unfortunately, we’re closer to building AI than knowing how to control it.”
Professor Tegmark hopes the summit will promote binding safety standards “as in all other important industries.”
