few-shot learning
The capability of learning a concept from only a handful of examples, as a toddler distinguishes a cat from a dog after about a dozen exposures; contrasted with current AI's need for massive datasets
As defined in Kevin Kelly on Advice, AI, and Technology
Other definitions
The ability of a machine learning model to learn new tasks or concepts from only a small number of examples without requiring extensive retraining
Metas AI Boss Says He DONE With LLMS...
Where it comes up · 5
- Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints - [Invest Like the Best, EP.469]2 claims
- 10 years of NLP history explained in 50 concepts | From Word2Vec, RNNs to GPT2 claims
- An audio version of my blog post, Thoughts on AI progress (Dec 2025)1 claim
- Gwern — Anonymous writer who predicted AI trajectory on $12K/year salary1 claim
- Kevin Kelly on Advice, AI, and Technology1 claim
Most important claims · 7 of 7Search all claims for “few-shot learning” →
- Although AI bulls fairly criticize bears for moving the goal posts, some goal-post shifting is justified: we keep solving what we thought were sufficient bottlenecks to AGI (general understanding, few-shot learning, reasoning) yet still lack AGI, so the rational response is to conclude that prior definitions of AGI were too narrow and there is more to intelligence and labor than previously realized.An audio version of my blog post, Thoughts on AI progress (Dec 2025)
- Current AI relies on neural nets needing billions of data points, unlike a toddler who learns cat-versus-dog from a dozen examples; once few-shot training arrives, training sets will be tiny and artists will clamor to be included to be counted among the few influential ones—making intellectual-property objections a distraction.Kevin Kelly on Advice, AI, and Technology
- GPT-3 was the crucial test of the scaling hypothesis: a massive scale-up from GPT-2 on a non-narrow task, where unimpressive results would have falsified scaling and impressive ones would confirm it; the few-shot learning chart on the first pages confirmed for Gwern that we live in the 'scaling world' where Legg, Moravec and Kurzweil were right.Gwern — Anonymous writer who predicted AI trajectory on $12K/year salary
- GPT-3 showed that large models can be few-shot or zero-shot learners, where few-shot learning refers to the ability where the model is shown some demonstrations or examples of a task during inference as conditioning.10 years of NLP history explained in 50 concepts | From Word2Vec, RNNs to GPT
- GPT-3 demonstrated one-shot and zero-shot learning abilities, and unlike fine-tuning, few-shot learning does not update the network's weights but instead just conditions the model during inference with the prompt.10 years of NLP history explained in 50 concepts | From Word2Vec, RNNs to GPT
- Vision-language-action robot models are inefficient in data and cannot scale their data fast enough, so they are unlikely to be the approach that ultimately scales, and few-shot learning breakthroughs in robotics could come within six to 18 months.Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints - [Invest Like the Best, EP.469]
- Over the next six to 18 months, Patel expects breakthroughs in robotics that enable few-shot learning from a pre-trained robot model, making robots useful for niche tasks such as chalkboard cleaning rentals, which will be deflationary and keep token demand growing.Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints - [Invest Like the Best, EP.469]