NVIDIA GTC 2024 Recap
Published on
March 28, 2024
Authors
Ryan Richards, Shambhavi Srivastava
Share
Last week, the Appen team had the opportunity to attend the NVIDIA GTC event, a cornerstone gathering for professionals in AI, machine learning, and other cutting-edge technologies. GTC has always been a beacon for innovation, and for us at Appen, it represented a chance to dive deep into the trends shaping technology's future.
What excited us at GTC
Next-gen Blackwell GPU - 'world's most powerful chip' with 30x inference performance, 25x more efficient.
Project GR00T Robots - a general-purpose foundation model that promises to transform humanoid robot learning in simulation and the real world. Training robots with human imitation learning is very promising.
Launch of a software stack called NVIDIA Inference Microservices (NIM) - a containerized inference microservice with optimized inference engines to simplify LLM deployments.
GTC Panels Highlights
Nvidia CEO Jensen Huang discussed the origins and impact of transformer neural network architecture with its creators and industry pioneers. He hosted a panel with all eight authors of the legendary 2017 paper "Attention Is All You Need" which introduced the concept of transformers. Appen's early adoption of transformer model technology has enhanced our ability to develop state-of-the-art in-house Speech-to-Text models, showcasing our commitment to advancing AI automation.
Impactful GTC sessions that align with Appen's work Strategies for Safer AI (LinkedIn Post)
The "Strategies for Safer AI" session provided a collaborative forum for leaders from NVIDIA, ActiveFence, Cohere, and the Brookings Institution to discuss the imperative of safe AI practices. They conveyed a unified stance on the significance of safeguarding AI, aligning perfectly with Appen's ethos of ethical data annotation to ensure AI operates within the bounds of trust and responsibility.
Navigating the Large Language Models Frontier: Practical Strategies for Building Enterprise Applications Powered by LLMs (LinkedIn Post)
The panel on "Navigating the Large Language Models Frontier" brought together voices from LangChain, LlamaIndex, Glean, and NVIDIA. They delved into leveraging LLMs for enterprise solutions, highlighting challenges such as behavior management and domain-specific adaptation. This discussion echoed Appen's focus on providing the nuanced data these sophisticated models require to function optimally in diverse environments.
Navigating the Opportunity for Generative AI in Financial Services
In the "Navigating the Opportunity for Generative AI in Financial Services" session, experts from NVIDIA, Mastercard, Capital One, and Goldman Sachs discussed the cutting-edge applications of generative AI. Their conversation about ethical use and the potential for innovation within financial services paralleled Appen's commitment to supplying the high-quality data essential for AI to navigate the complex financial landscape successfully.
Appen also made an impact from our booth in the expo hall, engaging with industry leaders, including our current partners. Our booth was a focal point for discussing the crucial role of high-quality training data in AI success. Through these conversations, we reinforced our commitment to AI development and strengthened our industry connections.
Discussions around AI Bias were particularly prominent, shedding light on the pervasive issue of bias in AI systems. This theme resonates with our commitment at Appen to develop AI responsibly. NVIDIA's focus on creating equitable AI models is a crucial reminder of our duty to mitigate bias through diverse data sets and ethical practices. It reinforces our dedication to ensuring our AI technologies are not only innovative but also inclusive and fair.
The bridge between AI research and its practical implementation was another key theme that emerged from the conference. NVIDIA's demonstration of applying cutting-edge AI technologies across various sectors underscored the significance of translating research into real-world solutions. This theme aligns with our approach at Appen, ensuring AI technologies are not only advanced but also applicable and impactful in real-world scenarios. The true value of AI lies in its ability to solve tangible problems, guiding our efforts to make AI practical and accessible through human involvement.
The pace of innovation in the tech world is rapid and requires a continuous need for collaboration and knowledge sharing. We're already looking forward to next year's event, eager to explore how the insights gained will shape the future of technology.
(20min delay)
