Amazon Unveils Nova, a New Generation of State-of-the-Art Foundation Models
Amazon’s re:Invent conference this year brought new chips and data centers, but it also introduced its first frontier models in generative artificial intelligence, competing with OpenAI and Google.
Introducing Nova and Reel
Amazon debuted Nova, a "new generation of state-of-the-art foundation models that deliver frontier intelligence and industry-leading price performance." The models come in flavors suited to video generation (akin to OpenAI’s Sora) and image generation, which has become standard fare for large language models that integrate text and images. The models come with snappy names, too: "Reel" is the name of the video-generation model, and "Canvas" is the name of the image-generation flavor.
Demonstrations and Features
Reel, a video-generation model, generated a video using the keyword "A snowman in a Venetian gondola ride, 4k, high resolution." Canvas, the image-generation model, created a photorealistic image of an interior with the prompt, "A very fancy French restaurant."
How it Works
Nova makes extensive use of the retrieval-augmented-generation (RAG) approach to tap into databases, as well as "chain of thought," a process for producing output that is treated as a kind of reasoning exercise by the AI model. The training data to build the models is not disclosed, with Amazon stating only that the models were trained on data from a variety of sources, including licensed data, proprietary data, open-source datasets, and publicly available data where appropriate.
Responsible AI
The most remarkable part of the work is the extensive discussion of "responsible AI" — that is, avoiding things such as adversarial attacks on AI models by malicious threat actors. Amazon’s engineers made extensive use of so-called red teaming, where they sought to break the models by creating various kinds of attacks such as "prompt injection," crafting a language model’s prompt with keywords or phrases that would encourage the model to break its guardrails.
Conclusion
Amazon’s Nova models are an exciting development in the world of generative AI, but the details are hidden behind intellectual property safeguards. It remains to be seen whether Amazon has broken ground in the reliability and safety testing of Gen AI. The field will have to wait until the proper evaluations — benchmarks, metrics, etc. — to compare Amazon’s red-teaming against the competing methods out there, both open and closed-source.
FAQs
Q: What is Amazon’s Nova?
A: Nova is a new generation of state-of-the-art foundation models that deliver frontier intelligence and industry-leading price performance.
Q: What are the features of Amazon’s Nova?
A: Nova models come in flavors suited to video generation (Reel) and image generation (Canvas).
Q: How does Nova work?
A: Nova makes extensive use of the retrieval-augmented-generation (RAG) approach to tap into databases, as well as "chain of thought," a process for producing output that is treated as a kind of reasoning exercise by the AI model.
Q: What is responsible AI?
A: Responsible AI refers to the avoidance of things such as adversarial attacks on AI models by malicious threat actors.
