According to new data

Hi, Beatrice Nolan here, filling the artificial intelligence correspondent Sharon Goldman, who is outside today. For all the noise of companies and Silicon Valley hand, New research indicates that the US job market has not yet tested the end of the world of artificial intelligence that some have warned.

in New reportYale and BROKINGS budget laboratories said that they did not find any evidence of any “clear disturbance” of jobs since the ChatGPt launched from Openai in November 2022. The study found that most of the continuous transformations in the American professional mix, a scale of functions that people maintain, were already ongoing in 2021, and modern changes that do not appear more.

The researchers wrote in the report: “Although the professional mix changes more quickly than it was in the past, it is not a big difference and precedes the extensive introduction to the spontaneous organization in the workforce.” “At the present time, the measures of exposure, automation and increase do not appear any sign that they are associated with changes in employment or unemployment.”

Industries with high exposure to intelligence, such as information, financial activities, and professional and commercial services, have witnessed some descending transformations, but these trends have started to a large extent. before Chatgpt launch.

The conclusion is not completely horrific, although it fly in the face of some of the most dramatic claims of artificial intelligence. Historically, the main turmoil in the workplace has been revealed for decades, not months or years. Computers, for example, were not common in offices until nearly 10 years after they appeared for the first time, and it was longer before reshaping the workflow tasks. If it ends up to transform the labor market significantly as computers – or more, it is reasonable to expect that the wide effects will take more than three years.

Some executives also told me that they are taking an “waiting and seeing” approach to employment while assessing whether technology could fulfill the promises of productivity. This approach can slow employment and make the labor market feel slowing, but this does not necessarily mean that workers enjoy automation of their jobs.

While anxiety about the effects of artificial intelligence on the labor market today may be widespread, new data indicates that this anxiety is still very speculative.

Entering employment problems

Real employment pain was felt by university graduates and workers in beginners.

No one denies that artificial intelligence is better in the tasks of this group of workers, and that companies say increasingly A calm part loudly When it comes to beginners roles. But the allegations that artificial intelligence keeps new graduates out of work, not supported by the entire new data. When the researchers compared the unemployed rates for new graduates with those with more experience, new graduates seem to have slightly more difficult roles, but the gap was not large enough to suggest technology is the main factor.

The researchers found a slight increase in professional difference compared to the older graduates, which could reflect the effects of early artificial intelligence but can also be easily attributed to labor market trends, including reactions to employers and job requests for noise on replacement of workers. The report indicates that beginners’ conflicts are likely to be part of the broader labor market dynamics instead of a direct result of the adoption of artificial intelligence.

Recently, there have been also anecdotal signs of life in the labor market for beginners. For example, Shopify and Cloudflare increases the trainee this year, with Cloudflare Calling artificial intelligence tools in a way “To strike how new appointments in the team” instead of a substitute for new appointments themselves can contribute. Younger workers are usually more accepted, more passionate to experience, and more creative when it comes to the use of emerging technology, which can give companies that employ a feature. UK -based program Simon Wilison put it“The armed trainee with artificial intelligence tools can produce a much faster value than the trainees in previous years.”

The researchers have warned that the analysis is not a prediction, and that they are planning to continue to update their results. They also warned that the sample size is small.

Only because artificial intelligence did not significantly affect the labor market yet It does not mean that it will not be in the future. Some modern assessments, such as The new GDP from OpenaiIt has been shown that the leading AI models improve in the performance of professional tasks at the level of human experts or approximately half of the cases, depending on the sector. With artificial intelligence tools improved and companies improve their integration, technology can have a more direct impact on the workforce.

But should we think about Amnesty International as just the next computer, or a new industrial revolution? At least now, the jury is still abroad.

However, here are the rest of the news of artificial intelligence.

Beatrice Nolan
[email protected]
Beafreyanolan

Our wealth

We are not in the “Intelligence Winter” – but here how do we survive the cold surprise Sharon Goldman

The California Governor signs the safety law of artificial intelligence, forcing major technology companies to detect protocols and protect violations Publishing Nolan

How can the latest Openai and Stripe move to shop online as we know it Sharon Goldman

“You cannot cancel the subscription – Fa Ruburg

Amnesty International in News

Meta plans to use AI Chatbot conversations to customize ads. Meta will start using chats with AI’s assistant to form ads and content recommendations via Facebook and Instagram. The company announced the update of its recommendation system on Wednesday, adding that it will be valid on December 16, with the user signs that start on October 7. I told the company Wall Street Journal He will not use talks on religion, politics, sexual tendency, health, race and race to customize ads or content. This step will link the huge Meta investments in AI Tawylidi in its basic advertising works. Users cannot cancel the subscription, but those who do not use Meta AI will not be affected, According to magazine.

The Mira Murate’s Thinking Machines Laboratory launches its first product. Thinking machines, a laboratory of artificial intelligence by Openai CTO Mira Murati, launched a tool that automates the creation of dedicated AI models. Moratti said Wireless The tool, called Tinker, will help “enable researchers and developers to try models and make border capabilities within the reach of all people.” The team believes that giving users the tools needed to control the border models will lead to the removal of mystery from the form of controlling the models, and makes the advanced artificial intelligence available beyond large laboratories, and helps to open specialized capabilities in areas such as mathematics, law or medicine. The startup raised $ 2 billion in seed financing in July 2025, before the issuance of any products, and it consists of a team of senior researchers including John Shulemann, who established Openai and led ChatGPT. Read more than Wireless.

Openai’s launch of the new version of Sora. Openai Sora 2, Video and AI launched the next generation of AI, along with an accompanying application that allows users to create, share videos and revitalize AI. The new model improves the realistic movement, generates speech, provides a “veil”, allowing users to introduce themselves into videos by recording a short verification. However, according to Wall Street JournalThe new video generator requires copyrights to cancel the subscription. This means that the film studios and other IP owners should actively request that Openai exclude its copyrights from copyright from the videos created by the new version of Sora. A subsequent report of 404 media found that users are able to generate strange content and often include copyright -protected figures such as Pikachu and SpongeBob SquarePants and numbers from Simpson family. Read more 404 media here.

A new starter assembles Amnesty International researchers. The periodic company LABS, a new startup in San Francisco, which was founded by the participating institution in ChatGPT, Liam Fidus and former Deep Mend scientist Ecin Dujus Copuk, employed a series of best researchers of artificial intelligence from Openaii, Google Deepmind and Meta, according to the The New York Times. More than 20 researchers, including Rishab Agarawal, who was Boiled by the definition From Deepmind just a few months ago, they left their work in the main AI companies to join the startup that focuses on building artificial intelligence that accelerates the discovery of the real world in physics, chemistry and material science. Eastern $ 300 million in financing and plans to use robots to operate large -scale laboratory experiments. Read more from New York Times.

You have a calendar

6-10 October: world Amnesty International The week, Amsterdam.

October 21-22: Teddy San Francisco.

10-13 November: The top of the web, Lisbon.

November 26-27: The world of artificial intelligence, London.

2-7 December: NeuPIPs, San Diego.

December 8-9: Wealth storm Amnesty International San Francisco. Apply to attendance here.

Eye on artificial intelligence numbers

72 %

This is the percentage of the S&P 500 companies that have revealed risks related to Amnesty International this year, according to the Conference Council, a non -profit institution for research and business membership, and Esgauge, a data analysis company. The general company has increased the detection of artificial intelligence as material risks in the past two years, as the share of the S&P 500 companies under the risk of the prosecution was martyred from 12 % in 2023 to 72 % this year.

The reputable risks are the most male about artificial intelligence, which were unveiled by 38 % of companies in 2025. Cyber ​​security was again close, was killed by 20 % of companies in 2024 and 2025. While all sectors reveal risks, financial care, health care, and witnessed the smallest rise. This may be because financial and health care companies face organizational risks related to sensitive and fair data, while industries largely work to expand the scope of automated operation and robots.

“The high disclosure of the risk related to the prosecution reflects the rapid circulation of the aerobic organization through the functions of companies in recent years, as companies included deeper in areas such as supply chains, customer participation, and product development,” said Andrew Jones, the main researcher in the conference board, said luck. “With the expansion of adoption, companies have increased their internal focus on governance, compliance and operational considerations, with panels, risk committees and legal teams that evaluate the potential challenges of data privacy and bias to organizational uncertainty and responsibility.”

The dramatic increase in disclosure indicates that more companies see the integration of artificial intelligence as a material danger that must be managed and connected to investors. The results were based on the 10-K files from the S&P 500 companies available until August 15, 2025.