percy liangThe director of the new Stanford research center says he has heard criticism but believes some may have misunderstood the goal of the project.
Liang says that large machine learning models that are referred to as “foundational” appear unique and important because of their ability to handle real-world complexity, as demonstrated by the capabilities of large language models. He says the feedback is part of a healthy academic debate. “All these criticisms are welcome,” he says.
Liang says the Stanford researchers are acutely aware of the limitations of these models and describe some in their research paper. Nor do they believe these models are needed for further leaps in AI, he says.
“It’s just kind of unbridled raw potential,” Liang says, “that we need to find a way to harness and contain.”
- The latest on tech, science and more: Receive our newsletter!
- Can robots evolve into machines of loving grace?
- 3D printing helps ultracold quantum experimentsUltracold Quantum ExperimentUltracold Quantum ExperimentUltracold Quantum Experiment go small
- How community pharmacies stepped up during Covidpharmacies step up during covidpharmacies step up during covidpharmacies step up during covid
- The Artful EscapeThe Artful EscapeThe Artful EscapeThe Artful Escape is psychedelic perfection
- How to send messages that disappear on their ownMessages that automatically disappearMessages that automatically disappearMessages that automatically disappear
- our new databaseour new databaseour new database
- Nerdshala GAMES: Latest Tips, Reviews & M. get receiveTips, Reviews, & MTips, Reviews, & MTips, Reviews, & More
- Torn among the latest phones? Never fear – check us out iPhone Buying Guide And favorite android phone