Google's recent announcement about the deprecation of Vertex AI AutoML text models marks a significant change in the AI landscape. Starting September 15, 2024, users will no longer be able to train or update models for text classification, entity extraction, and sentiment analysis using Vertex AI AutoML. Instead, Google is pushing users towards Vertex AI Gemini, which promises enhanced user experiences through improved prompting capabilities. Existing AutoML text models will remain available until June 15, 2025, marking a pivotal moment for how businesses and developers will utilize Google's AI tools for natural language processing (NLP) tasks.
Since the rise of ChatGPT, Google has increasingly prioritized Generative Artificial Intelligence (Gen AI), as demonstrated by their own GenAI models, including Gemini. Recently, Gemini has undergone significant enhancements, such as an increased token limit to 1 million, improved handling of multimodal data, and a strong push for GenAI applications like the Gen App Builder. Notably, Gemini Pro has become the leading GenAI model within the Vertex AI ecosystem.
But what does this shift mean for Google's previous services, such as Vertex AI AutoML for text? It appears that these “simpler” tasks can now be managed by Gemini. While this may be a forward-looking move, it poses immediate challenges for small to medium-sized companies that have built their machine learning infrastructure around Vertex AI AutoML services. The short timeline for this transition feels abrupt, and businesses relying on these services may face significant disruptions.
Gemini Pro has only been available since December 2023. Adopting and integrating new technologies takes time—testing the model, debugging, and creating comprehensive documentation and tutorials are all essential steps. The sudden removal of AutoML services in favor of Gemini could be premature for many companies, especially those that need to retrain their models regularly.
If your company falls into this category (small to medium-sized businesses relying heavily on Vertex AI AutoML), you may be facing an internal scramble to adapt. It's crucial to establish a migration path that ensures continuity in serving your customers. Here are some potential strategies: Possible make visualization for this or roadmaps
In upcoming blog posts, we will delve deeper into each of these migration paths to help you navigate this transition effectively. If your company has any questions and concerns, please reach out.
