Jacob Coxon, the Anthropic researcher who recently resigned with his warnings becoming viral, spoke to BBC On the other hand, there is a need for greater transparency in the implementation of Community legislation. Artificial Intelligence (AI).
Some persons in the industry have suggested that comments on the risks and strength of Artificial Intelligence may be aimed at creating above «noise» for her
The people who deal with this technology are «Really scared.» by the pace of its evolution and by what this could mean for humanity.
«I believe that if we do not slow down the current pace of progress, there is a good chance that we will all die in the near future.», said features.
Former head of the 27-year-old, the director of Anthropic Dario Amondei, also recently requested that the development of Artificial Intelligence be delayed — Although some have questioned the motives behind this move.
The heads of two competitive Artificial Intelligence companies, Sam Altman of OpenAI and Elon Musk of xAI, both said they agree with Amondei's proposal to slow down and regulate at branch level, as well as for independent monitoring of the development of Artificial Intelligence models.
Amonday wrote in an article on Saturday that the development of technology is not contested, but the risks associated with it are «serious» and that companies and governments must be given time to deal with them.
Coxon – who worked at OpenAI before joining Anthropic – welcomed the proposal to slow down, but said it should be coordinated with China in order to avoid «a race on an international scale».
«People who work in these companies are asking for regulation with total seriousness because they feel trapped in a race. And they fear the consequences of this race», stated on the show «Sunday with Laura Kuenssberg».
The statements of former Anthropic researcher:
How would you feel about a «Disclosure» AI?
The most difficult question to answer, according to Coxon, was what a «Disclosure» of Artificial Intelligence.
One of the dangers noted in Amonday's statements was the possibility that a cluster of bots would act as a supercomputer and take control of the internet.
Coxon said this scenario could become a reality within six months to a year.
In response to Coxon's departure, an Anthropic spokesman told BBC News: «We have always been clear that Artificial Intelligence will bring both enormous benefits and unprecedented risks.».
«To address these risks, we continue to develop models with some of the strongest protection measures in the industry», note.
The company has pioneered the study of how Artificial Intelligence models operate, the spokesman added.
It was the first to publish a framework for mitigating the risks involved in their development. — and also tries «intensively» its models and publishes the findings in order to facilitate control and prevent incidents «the mismatch of Artificial Intelligence».
«This work is also the reason why we believe it would be for the benefit of the world if the industry adopted a legitimate and verifiable way of working together to regulate the pace at which we circulate strong models», they stated.
What Coxon's colleagues fear
Coxon said his colleagues fear that the risk may even occur within the next two years.
He said the staff of the Artificial Intelligence companies «plans what he'll do with his life and thinks about the implications of his work», while some «consider buying land somewhere, because they fear so much the instability that the rapid progress of Artificial Intelligence will cause».
«All of them have this in mind every day while working on this technology», add.
However, Cox showed some optimism about the future of Artificial Intelligence, stating to the BBC that people investigating this technology «They sincerely want to see the positive results» in sectors such as «treating diseases and improving everyone's life».
The debate on the risks involved in the unlimited development of AI has been rekindled following Jacob Coxon's statements.
«We really and honestly believe that Artificial Intelligence could kill all people!»
Many others in the industry have also expressed their concerns after Coxon's posting on social media, including Anthropic's scientist, Ivan Hubinger.
«We really and honestly believe that Artificial Intelligence could kill all people! Personally, I appreciate that this possibility is more than 10% within the next decade.», said Hubinger.
Computer scientist and Nobel Prize winner Jeffrey Hinton, known as «Godfather of TN», stated on Friday at the BBC that the probability of 10% AI eliminating all people «It's not absurd.».
Meanwhile, Mark Warner, CEO of the Faculty company specializing in the safety of Artificial Intelligence, told the show «The World This Weekend» of the BBC that is «extremely difficult to calculate probability» AI to eliminate all people.
«But it is important to recognise that these people are completely honest about what they say.», he added, pointing out that the risks associated with Technological Intelligence have been identified for years by the heads of several companies in the industry.
Former Prime Minister of the United Kingdom, Risi Sunak, who is a hired advisor to Anthropic, wrote in the Sunday Times that he himself is concerned about the danger that Artificial Intelligence poses to humanity, despite generally maintaining an optimistic stance.
Artificial interest?
However, some persons in the industry have suggested that comments on the risks and strength of AI may aim to create above «noise» For her.
The CEO of the Hugging Face Artificial Intelligence platform, Cleman Delang, noted on social networks: «Sorry, but asking Jacob (p. Coxon) about the risk of extinction due to Artificial Intelligence is like asking your air conditioning technician about climate change. I am not saying that they are necessarily indifferent or wrong in themselves, but let us see things in their right dimension».
However, after the publication of Amondei's article, Delang offered to contribute to the possible solutions proposed by Anthropic's chief.
Meanwhile, Nvidia's head, Jensen Huang also commented on Cox's statements at a conference organized by investment bank Goldman Sachs last week, as reported to the BBC by several people participating in the event. They reported that he rejected them as untrue.
Huang had previously stated that the idea that Artificial Intelligence «It would mean the end of humanity.» was «absolute nonsense».
And while it may have a business interest in the flourishing of Artificial Intelligence —Nvidia builds chips that feed intelligence probe systems—, his comments reflect a growing reaction to Silicon Valley against existing warnings from current and former employees.
Other critics argue that Anthropic is trying to provoke a regulatory initiative aimed at preventing competition, so that it remains a monopoly with two «players».
According to BBC, Anthropic is preparing for an initial public stock offer on the stock market, which may break every record, allowing the public to purchase shares of the company.
OpenAI, whose value was recently estimated at $852 billion, was expected to follow the same course, but Altman said Friday that this would not happen this year, citing security reasons.
