Get ready for a game-changer in the world of AI! AWS has just unveiled a suite of cutting-edge AI models and a groundbreaking service that empowers businesses to take control of their AI journey.
The AI Revolution is Here!
Amazon Web Services (AWS) has taken the tech world by storm with its latest announcement at AWS re:Invent. Led by CEO Matt Garman, AWS has introduced Nova 2, an impressive fleet of four new AI models, building upon the success of the original Nova family.
But here's where it gets exciting: AWS isn't stopping at just releasing new models. They're also launching a service called Nova Forge, which allows enterprise customers to create their own custom AI models, known as Novellas, for an annual fee of $100,000.
The first version of Nova was unveiled last year, and it quickly gained traction. AWS released four text-generating models and one image-generating model, which have since been used by tens of thousands of customers, ranging from marketing giants to innovative startups.
This year, AWS is taking Nova to the next level with some powerful upgrades. Let's dive into the details:
- Nova 2 Lite: A cost-effective reasoning model that 'thinks' before it responds. It can process text, images, and videos, making it perfect for everyday tasks.
- Nova 2 Pro: A reasoning agent designed for highly complex tasks like coding. It can handle text, images, videos, and speech, making it a powerful tool for tech leaders.
- Nova 2 Sonic: A speech-to-speech model for conversational AI, adding a human-like touch to your interactions.
- Nova 2 Omni: A multimodal reasoning and generation model that can process and generate text and images from various inputs, including images, text, video, and speech.
And this is the part most people miss: the controversy surrounding AI customization. AWS's Nova Forge service allows enterprises to access and train pre-existing models on their proprietary data. Garman believes this solves the issue of models 'forgetting' core reasoning when customized with new data. He compares it to humans learning languages, suggesting that model training is a similar challenge.
However, this interpretation is not without its critics. Some argue that the 'forgetting' phenomenon is a natural part of AI evolution and that attempting to prevent it may limit the model's ability to adapt and learn.
What's your take on this? Do you think AWS's approach will revolutionize AI customization, or is it a step too far? Share your thoughts in the comments below!
Early adopters of Nova Forge include Reddit, Sony, and Booking.com, so keep an eye out for their innovative AI implementations.
For more insights and coverage of the latest in enterprise tech, stay tuned to TechCrunch's event in San Francisco from October 13-15, 2026.
This article was written by Becca, a senior writer at TechCrunch, who has previously covered venture capital trends and startups for Forbes and the Venture Capital Journal. You can reach out to Becca or verify her outreach via email at rebecca.szkutak@techcrunch.com.
