Thomas Wolf’s Thomas Wolf said that Amnesty International does not create new knowledge
- Thomas Wolf said artificial intelligence excels in following the instructions, but it is struggling to create new knowledge.
- Artificial intelligence needs to question its training data and take non -intuitive methods, as The Luging Face Exec wrote on X.
- WOLF comments come as Tech focuses on Agency AI.
Thomas Wolf said artificial intelligence outperforms the following guidelines – but it does not pay the limits of knowledge.
Amazon and Nvidia analyzed the boundaries of the large language models in Thursday’s publication on X. and NVIDIA, the main director of science and founder of Huging Face, an Amnesty International Open source company supported by Amazon and NVIDIA, in a publication on Thursday on X.
He wrote that the field produces “excessively compatible assistants” instead of the revolutionaries.
Now, Amnesty International does not create new knowledge, Wolf books. Instead, it is just filling the voids between the current facts – what it called “filling the horns”.
Wolf argues that in order to lead artificial intelligence real scientific breakthroughs, it needs to do more than recovering and synthesizing information. Artificial intelligence must question its training data, take non -intuitive methods, create new ideas from the minimum inputs, and ask unexpected questions that open new research paths.
Wolf also listened to the idea of ”compressed twenty -first century” – a concept of October article by the CEO of Anthropor, Dario Ameudi, “The Grace Loving Machine”. Amodei has written that artificial intelligence can accelerate the scientific progress that the expected discoveries over the next hundred years can occur in only five to 10.
“I read this article twice. The first time I was completely surprised: Amnesty International will change everything in science within five years, I thought!” Wolf wrote on X.
Unless artificial intelligence research was made gears, Wolf caution, we will not get a new Albert Einstein at the Data Center-once a future full of “Yes-Men on Servs”.
The wolf did not respond to a request for comment, he was sent outside the standard working hours.
The climb of the agent artificial intelligence
Wolf’s comments come at a time when the world of artificial intelligence focuses on AIC.
Sam German, CEO of Openai, expected this to be the first year “agents” – a set of artificial intelligence tools that can independently perform tasks – “join the workforce.”
“If 2024 is the year of LLMS, then we believe that the year 2025 will be the year of the agent of artificial intelligence,” Braved Akirajo told Business Insider in January. He is an administrative manager at Insight Partners, a VC company that includes the plays of the author, Jasper, and Torq.
Investors are betting on this idea. According to Pitchbook data, startups that explore the agents’ application 8.2 billion dollars last year.
Unlike artificial intelligence assistants, who mainly recover and summarize information, agents can break up complex tasks, make decisions, and improve their approach based on results.
The researchers also used Amnesty International to achieve scientific breakthroughs.
Professor Oxford Matthew Higgins, ALPHAFOLED2, an Amnesty International tool from the alphabet, used to break the main malaria protein-which has been struggling with his laboratory for years. This penetration led to the test of the experimental malaria vaccine in people.
“Without alphafold,” he said to Bi in 2023.