Hallucination is not a solved problem
6 December 2023 (MIT Technology Review). Sundar Pichai, CEO of Google, speaking to MIT Technology Review's Mat Honan at Google's offices in Mountain View, California, on the eve of Gemini's launch, December 2023, about Gemini and the limits of large language models.
Google CEO Sundar Pichai on Gemini and the coming age of AI
More from Sundar Pichai · AI and safety · AI and science · 2023
Who else?
What it refers to: AI models producing hallucinations or revealing private training data, which the answer calls active fields of research with more work to be done.
Other voices
- “They literally don't know the difference between truth and falsehood” Gary Marcus, Researcher,
- “Wrong or biassed but sounded right and authoritative, i think it's dangerous” Michael Black,
- “On the highway towards Human-Level AI, Large Language Model is an off-ramp” Yann LeCun, Researcher,
- “Embers of autoregression” paper by McCoy, Yao, Friedman, Griffiths,
- “We're hitting the upper limits of the LLMs right now” Marc Benioff, Builder,










