Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Ask HN: Pull the curtain back on Nvidia's CES keynote please

Jan 12, 2025 - news.ycombinator.com
The author, with 15 years of experience in robotics and software engineering, has traditionally viewed AI and large language models (LLMs) as tools for augmenting specific tasks rather than transforming engineering fundamentally. However, NVIDIA's CES keynote challenged this perspective by suggesting that AI can solve any problem across modalities and that future engineering will focus more on data gathering and AI integration rather than first-principles problem-solving. Despite the compelling vision presented, the author's personal experience with AI/LLMs has been less revolutionary, noting their usefulness in suggestive tasks and prototyping but highlighting issues like unreliability for authoritative answers, the need for major rewrites for production-ready code, and persistent hallucinations.

Seeking real-world insights beyond sales pitches, the author turns to the Hacker News community for firsthand experiences with NVIDIA's AI tools and LLMs/multi-modal AI. They are interested in understanding what these technologies are genuinely achieving, their limitations, reliability, and whether challenges are due to resource constraints or fundamental barriers. The author is particularly curious about whether these tools accelerate reaching production compared to traditional methods and seeks honest feedback from those actively working with these technologies.

Key takeaways:

  • The author has a traditional engineering background and views AI/LLMs as tools for augmenting tasks rather than reshaping engineering.
  • NVIDIA's CES keynote suggests AI can solve problems across modalities and shift engineering focus to data gathering and AI integration.
  • The author's consumer experience with AI/LLMs shows they are useful for suggestive tasks and prototyping but unreliable for authoritative answers and production-ready code.
  • The author seeks real-world insights on the practical use and limitations of NVIDIA's AI tools and LLMs from those actively working with these technologies.
View Full Article

Comments (0)

Be the first to comment!