Google Introduces PaLM API & MakerSuite; Simplifies Generative AI Development Lifecycle
Google Cloud is revolutionizing the way developers could be using generative AI for building new applications and platforms in the near future. The world’s biggest AI company announced a new DevOps platform specifically built for accelerating and simplifying the generative AI development lifecycle. It’s called PaLM API, a new AI developer offering to test and experiment with Google’s Large Language Models (LLMs) and generative AI tools. To make prototyping quick and more accessible, developers can integrate PaLM API with another newly-launched tool called MakerSuite. Both tools are currently in private preview mode and would be accessible soon.
This is what PaLM API brings to your generative AI development lifecycle if you are an app builder.
Hands-on access to Google’s LLMs
Google is continuously transforming the way AI developers use large language models for building applications. Since the launch of Google Transformer Architecture in 2017, the technology innovation company has introduced many large language models trained in massive troves of data such as GPT-3. These include GLaM, PaLM and LaMDA. With PaLM API, AI developers can gain entry into Google’s LLMs and optimize these for specific multi-turn use cases such as virtual assistants, search results, product recommendations, robotic automation and AIops. Google has confirmed PaLM API would add more capabilities and models with varying sizes very soon.
Accelerated AI development with PaLM + MakerSuite
Google’s intent behind launching PaLM with MakerSuite is straightforward — they want to not just simplify the entire Generative AI development lifecycle in one shot, but also create an agile environment where all the LLMs intersect and synthesize new models. Powered by MakerSuite capabilities, PaLM API would break down the barriers in AI generation and augment every dataset with new types of synthetic data in a customized way. DevOps teams can build their favorite applications using programming languages and frameworks of their choice. It supports open source libraries and downstream applications built using TensorFlow, Keras, Python and Node.js.
Build New AI Models with Out-of-Box Functionalities
Developers working with generative AI seek three things from their LLMs — accurate performance, sequential training even without access to direct data sets, and fewer parameters per task. To perform specialized tasks in advanced workflows, PaLM with MakerSuite would allow faster operations that lead to better and more accurate results. MakerSuite makes testing and iterations on your preferred tuned model easier and smoother. According to Google Cloud’s official statement, “High-quality data is crucial when developing with AI, and developers are often limited by the data they have. Our tooling will allow you to generate additional data based on a few examples, and then you’ll be able to manage and manipulate the data from there. This synthetic data can be used in various scenarios, such as tuning or evaluations.”
You will Own Your Data and Domain
Generative AI extracts lot of information from external sources. There is a perennial danger of data security and compliance when you are working with big data for AI applications. PaLM API ensures you are always working with your own data or external data sources in a safe and responsible environment guided by AI principles. If you are new to generative AI marketplace, starting with MakerSuite and PaLM API would give you the necessary control to define safety and responsibility without hampering your outcomes or contexts.
In addition to launching PaLM API and MakerSuite, Google also announced new upgrades to its Vertex AI offering, as well as doping its Google Workspace with advanced AI-powered experiences built on intuitive search and conversational AI.
Comments are closed.