Launched at the end of 2022, ChatGPT started creating a fever at the beginning of this year when it reached 100 million users in just two months. This AI and similar chatbot models from Google, Microsoft… become useful tools for users in both work and entertainment. However, the new trend also gives rise to debates related to the dangers of artificial intelligence and the responsible development of artificial intelligence.
More than 1,000 experts signed a letter calling for AI control
The breakneck pace of AI development has worried many top technology leaders. In March, billionaire Elon Musk , Steve Wozniak – co-founder of Apple, Emad Mostaque – CEO Stability AI, professor Yoshua Bengio – one of the founders of AI, and more than 1,000 researchers, leaders, experts Other experts have signed a letter calling for a halt to super AI training.
Elon Musk. Photo: Reuters
Authored by the Future of Life Institute, the letter calls on artificial intelligence labs globally to “stop training AI systems more powerful than GPT-4 for at least six months.” . This move is not to limit technology but is a necessary step back to control and prevent the risk of abusing AI for dangerous purposes.
Immediately after appearing, the letter quickly attracted attention with mixed opinions. Gary Marcus, a New York University professor, told Reuters that “the letter’s content is not perfect, but the spirit is correct”. On the contrary, some experts say the problem is exaggerated. According to professor Emily Bender at the University of Washington, the orientation letter was “in a very vague and ineffective way.” Arvind Narayanan, an associate professor of computer science at Princeton University, also assessed that the letter was full of exaggeration, making it “more difficult to address the real harm that is being done by AI”.
Meanwhile, despite actively calling for control of artificial intelligence, Musk still founded the xAI company in July and announced super AI Grok to compete with ChatGPT in November.
Fake photo of Trump arrested, Pentagon on fire
Besides the ChatGPT craze, the photo AI tool also became a controversial topic when it was able to create realistic photos just through a short text description.
In March, a user used the Midjourney tool to create a photo of former US president Donald Trump being arrested. According to the Washington Post, the photo was quickly shared widely on social networks and worryingly, many people believed it was real.
Photo of DonaldTrump being arrested was created by AI Midjourney.
By May, a fake photo of a burning Pentagon with a large column of smoke also spread on Facebook and Twitter, including a number of accounts with millions of followers, with blue ticks. The information then appeared on several major news sites. CNN and NBC News noted that the US stock market was even affected in a short time. The Dow Jones index fell 80 points, the S&P 500 index fell 0.3% to a session low and then recovered after the photo was verified as fake.
According to experts, easily creating fake photos using AI can cause serious social harm, contributing to the spread of false information. “As AI image generators become popular and sophisticated, it is not difficult to imagine that one day they can manipulate and spread false information about an event,” commented technology site TechCrunch .
Hollywood declares war on AI
In mid-July, hundreds of Hollywood actors, directors, and filmmakers took to the streets to protest the introduction of artificial AI into cinema. According to WSJ , this is considered the largest strike of workers in the American film capital in the past 60 years.
Actors, writers, screenwriters… protested in front of Warner Bros headquarters on July 13 about salary reduction and the threat caused by artificial intelligence. Photo: AFP
In Hollywood, studios see the potential of AI in production. These tools are now used to speed up the storyboarding process, create graphics to visualize how the story unfolds, and to design the set. The 2023 Oscar-winning film Everything Everywhere All At Once uses some scenes from AI, which also makes the filmmakers “both happy and worried”.
According to the Washington Post, some famous actors, including Keanu Reeves , believe that using AI for Hollywood movies is “scary” and could be a way for film project owners to unfairly pay artists. . Reeves assesses that AI can bring some benefits, but is more inclined to view this technology as a threat to creators.
In May, the Writers Guild of America (WGA) went on strike for nearly 5 months and asked studios not to use AI to write documents, nor to take writers’ content to train AI. The WGA has reached an agreement with the studios. Although this is considered an initial victory, experts believe that similar conflicts will soon continue considering the current pace of AI development.
Sam Altman was fired
OpenAI CEO becomes a prominent face of the technology field in general and AI in particular in 2023 thanks to the popularity of ChatGPT . However, on November 17, he was suddenly fired by OpenAI via Google Meet. Under pressure from investors, one day later, the board of directors invited him to negotiate but failed. By November 20, more than 700 company employees simultaneously signed a letter threatening to quit if Sam Altman did not return. On November 22, OpenAI negotiated for the second time, bringing him back to the CEO position.
OpenAI CEO Sam Altman in a testimony before the US Congress in May. Photo: Reuters
The OpenAI scandal is considered the most shocking event in the AI community this year. “Like an atomic bomb explosion,” Pietro Schirano, founder of AI startup AverArt, wrote on X. Meanwhile, Elon Musk said that there was ” something scary ” that motivated the OpenAI leadership Take down Altman.
To date, the actual cause of the dismissal has not been announced. However, according to Reuters , the OpenAI board of directors made a sudden decision after receiving a confidential letter , warning about the dangers and potential power of Q* (Q-Star) – an early form of artificial general intelligence ( AGI ), developed by OpenAI in parallel with ChatGPT.
Bao Lam
EN 
VI
Leave a Reply
You must be logged in to post a comment.