 |
VOOZH | about |
|
VOOZH | about |
We’re so glad you’re here. You can expect all the best TNS content to arrive Monday through Friday to keep you on top of the news and at the top of your game.
Check your inbox for a confirmation email where you can adjust your preferences and even join additional groups.
Follow TNS on your favorite social media networks.
Become a TNS follower on LinkedIn.
Check out the latest featured and trending stories while you wait for your first TNS newsletter.
Advances in transformer neural networks and generative artificial intelligence (AI) are driving one of the biggest technology shifts in modern times. They also have the potential to unlock innovation and ingenuity at scale.
As AI development evolves, data reigns supreme. It’s the lifeblood that fuels machine learning projects, transforming mere concepts into actionable insights. However, the path to effectively leveraging data in AI projects is laden with challenges that can hinder adoption and achieving transformative value.
To strengthen AI development, a paradigm shift is emerging with the transition from a model-centric to a data-centric approach to AI. This shift can significantly help with reducing hallucinations and biases in generative AI systems. Focusing on data-centric AI and bringing models closer to data will improve the output of AI models and enable businesses to unlock their full potential.
A model-centric AI approach is the traditional way machine learning evolved. It involves iterating to improve the performance of a model, aiming to produce the best model for a given data set. Researchers and engineers spend considerable time fine-tuning model parameters, layers and other architectural elements. However, the data set is often treated as secondary because, historically, building and fine-tuning models has been complex and resource-intensive, requiring deep expertise to generate meaningful outputs.
In contrast, a data-centric approach improves the quality of the data a model is trained on. It includes data cleaning, augmentation and ensuring the data is representative of the real-world scenarios where the model will be deployed.
As AI models mature, diversify and expand in complexity, organizations should work on enhancing data quality and forging a closer alliance between models and data. In this evolving narrative, the shift is necessary and clear: bring models closer to the data rather than shuttling data to the models. The result is an amplification in model output quality and a reduction in hallucinations that often plague AI systems. This data-centric AI approach is a cornerstone for organizations that want to deliver generative and predictive experiences rooted in the freshest data.
While data-centric AI is the path forward, model-centric AI still plays a crucial role. It’s particularly important in scenarios where data is limited or the goal is to explore the limits of model complexity and performance. It remains vital for pushing the frontiers of AI research and for solutions where high-quality data might not be readily available.
By shifting to a data-centric AI approach that ensures the quality and relevance of data, organizations can benefit in the following ways:
One of the quintessential advantages of a data-centric approach is the ability to deliver experiences that are closely aligned with real-world scenarios. Unlike the model-centric approach, where models often grapple with the fallacies of low-quality data, data-centric AI strives to bridge the chasm between AI models and the dynamic realities they aim to navigate.
AI hallucinations, characterized by the generation of incorrect or fabricated information, are primarily a symptom of flawed data. Pivoting to a data-centric approach enhances the likelihood of reducing these errors. Training models on cleaner, more representative data sets leads to outputs that are more accurate and reliable.
With a solid foundation of high-quality data, organizations can unlock the full spectrum of AI’s predictive and generative capabilities. This shift makes AI more capable of interpreting existing data patterns while also generating new insights and experiences, fostering a culture of innovation and informed decision-making.
Transitioning from a model-centric to a data-centric AI approach represents a fundamental shift in thinking. It’s about placing data at the core of AI’s transformative journey. The shift is more than a mere technical adjustment; it’s a conceptual realignment that places data at the heart of AI. As organizations tread this path, they must cultivate a robust data infrastructure, nurture data literacy and foster a culture that values data as the foundational cornerstone of AI’s promise.
Building robust AI solutions necessitates a nuanced understanding of when to emphasize data and focus on model innovation. A balanced approach that uses the strengths of both model-centric and data-centric AI is essential to addressing today’s AI challenges so that organizations can get the most value from their AI projects. To help ensure AI models are developed on the freshest data and are accurate and reliable, organizations must embrace the shift to data-centric AI.
