As we step into 2024, the AI landscape is poised for an unprecedented wave of power and transformation. An article from The Economist titled, “Generative AI Will Go Mainstream in 2024” postulates that 2024 will be the year enterprise adoption of the technology will truly take off. This is, of course, backed by the notion that many companies spent much of 2023 experimenting with it and are at the point where it’s time to move from theory and experimentation to more wide-scale implementation.
This year, instead of compiling a (potentially overwhelming) laundry list of possible trends that we believe will take precedence in 2024, we’ve taken a new approach: In addition to one core prediction (a “hot take,” if you will) from Dataiku’s co-founder and CEO Florian Douetteau, we’ve invited Oz Karan of Deloitte and Ahmad Khan of Snowflake to participate in our e-magazine and share their most pressing advice as organizations confront their analytics and AI strategies in the new year. This blog shares a high-level overview of each contribution, but not to fear — we’ll be sharing the full features on the blog in the new year!
A Coming Era of Compute Cost Volatility
Dataiku’s Florian Douetteau anticipates a new era of compute cost volatility, specifically revolving around Graphics Processing Units (GPUs), a cornerstone in the AI infrastructure. GPUs, known for their parallel processing capabilities, play a pivotal role in mastering Large Language Models (LLMs), which are the technical backbone of Generative AI.
Douetteau emphasizes that the demand for GPUs will surge, driven by the rising adoption of Generative AI in the enterprise. However, the supply of GPUs remains uncertain, intertwined with the dynamics of the semiconductor supply chain, susceptible to global events like natural disasters or geopolitical shifts.
In response to this emerging challenge, organizations must equip themselves with strategies to navigate compute cost volatility effectively. Douetteau outlines key considerations for businesses:
- Analyzing Tradeoffs: Companies need to assess the tradeoffs between cost and quality, ensuring efficient deployment of computational resources.
- Flexibility in Model and Service Providers: The ability to switch between models and service providers allows organizations to adapt their cost posture based on variable costs.
- Managing GPU Clusters: Companies may opt to manage their own GPU clusters, providing greater control over computing infrastructure and potential long-term cost benefits.
- Optimizing Models for Specific Use Cases: The adoption of technologies that optimize models for specific use cases becomes crucial in enhancing GPU utilization efficiency.
Dataiku, the platform for Everyday AI, emerges as a crucial ally in this era of compute cost volatility. The platform's robust orchestration across data science and ML workflows enables organizations to optimize resource utilization and reduce unnecessary compute costs. Features like scalable infrastructure, efficient data management, and collaborative environments empower businesses to navigate the challenges posed by GPU dynamics.
An Opportunity to Build Trust: Deloitte's Perspective
Deloitte, a stalwart in risk and financial advisory, brings a fresh perspective on the importance of trust in the age of AI (specifically, Generative AI). Oz Karan, Partner, Deloitte Risk & Financial Advisory, emphasizes the critical role of trust in the adoption of AI technologies. As AI becomes ingrained in everyday life, the distinction between trust in an organization and trust in its use of AI blurs.
The age of AI is upon us with the dawn of a broad spectrum of Generative AI capabilities available to consumers. But much like the technologies themselves, trust in these paradigm-shifting technologies will not be built in a day.
- Oz Karan, Deloitte
Deloitte's Trustworthy AITM Framework provides a comprehensive approach to help organizations build trust in order to maximize Generative AI’s business potential. The framework addresses the erosion of trust possibilities, understanding that AI is not infallible and requires responsible planning and response. Building AI solutions with trust by design becomes imperative, and Deloitte offers insights into regulatory guidance and AI Governance practices.
The Future of Advanced AI Is Simple: Snowflake's Approach
Snowflake, represented in the e-magazine by Ahmad Khan, Head of AI/ML Strategy, brings a refreshing perspective on simplicity in the era of advanced AI. In the age of Generative AI, the real differentiator lies in proprietary data and customized models. Snowflake's approach focuses on minimizing data movement, enabling developers to conduct advanced processing where data is already curated and governed.
Getting the most value from Generative AI will require organizations to define a holistic strategy that first establishes a robust data and model governance and then enables developers to accelerate LLM app development and analysts to leverage AI as part of everyday analytics.
-Ahmad Khan, Snowflake
Snowflake offers scalable infrastructure and LLM application stack primitives, such as Snowflake Cortex and Snowpark Container Services, allowing developers to build apps without unnecessary data movement. The emphasis on natural language interfaces and pre-built UIs, like Document AI, signifies Snowflake's commitment to making advanced AI accessible beyond technical experts.
Navigating the AI Waters With Confidence in 2024
As we embark on the journey through 2024, the collective insights from Dataiku, Deloitte, and Snowflake paint a comprehensive picture of the challenges and opportunities in the AI landscape. Compute cost volatility, trust-building, and simplicity emerge as the defining themes, requiring organizations to adapt, strategize, and collaborate with trusted partners like Dataiku, Deloitte, and Snowflake. In this era of unprecedented AI power, businesses equipped with foresight, agility, and frameworks for safe scaling will navigate the AI waters with confidence and emerge as pioneers in the transformative wave of technology.