This page was machine-translated and may differ from the original. View original
78% of U.S. Say “ChatGPT Will Contribute to Advancements in Science and Medicine”
ChatGPT-4 Released After 4 Months… Hallucination Errors ↓
Expectations of spreading demand for mega-scale AI robotic surgery, etc.
Expectations of spreading demand for mega-scale AI robotic surgery, etc.
With the emergence of ChatGPT causing a huge social stir in the industry, OpenAI's ChatGPT is recognized as contributing to the advancement of science and medicine.
OpenAI released a new multimodal language model, ChatGPT-4, just four months after publicly unveiling ChatGPT version 3.5 on the 14th, which shocked the world. ChatGPT-4 is currently available for use after upgrading for a fee of $200.
The newly unveiled ChatGPT-4 features multimodal capabilities to recognize photos and handwriting, and its Korean language performance has also been enhanced. It has demonstrated aspects of being "closer to humans," ranking in the top 10% of standardized tests such as the bar exam.
Early versions of ChatGPT faced backlash for producing a series of errors, nonsensical answers, and plausible lies. However, ChatGPT-4 stated that it improved AI accuracy by minimizing the creation of 'hallucinations'.
After ChatGPT was first launched last December, the Pew Research Center conducted a survey of approximately 11,000 Americans who had encountered it, asking for their opinions on how AI will impact the future.
44% of respondents answered that AI "will have an impact" on whether it can reduce the time reporters spend writing news articles. 11% said they were unsure.
On the other hand, the majority answered that AI "could provide significant advancements in scientific fields, namely medicine and weather forecasting." In particular, 78% of participants expressed the opinion that AI technology could impact robots performing surgery. Examples cited included predicting the protein structure of cells and producing drought and heat-resistant crops.
The industry believes that ChatGPT-4 will have fewer errors than previous versions. OpenAI CEO Sam Altman said, “ChatGPT-4 still has flaws and limitations, but it is amazing, more creative than previous models, and has fewer hallucination and bias errors.”
With the emergence of various mega-AI models predicted, diverse discussions regarding their socio-economic aspects are expected to continue.
본 기사에 대한 정정·반론·추후보도 청구는 보도 청구 안내를, 그간 게재된 보도문은 정정·반론보도 모아보기를 참고해 주세요.














