Rob Durst’s Post

View profile for Rob Durst, graphic

CEO & Co-Founder at Silver Bay Technologies

Just a thought…. WR2 AI, LLMs, and AI ‘agents’ - Apple’s Siri was formally incorporated as a ‘free’ feature in iPhone’s in 2010. That means Apple has access to 13 years of worldwide, multi-lingual usage data across all of it’s Siri enabled platforms for training. As I understand it, one restriction of LLM’s is access to large and varied training sets unencumbered by IP (like the legal complications that are now arising from ‘scraping the web’). You can always ‘buy’ one or more programs to sift through the training sets, and in fact this may be a way to monitize them based on usage. A ‘vetted’ data set may also reduce or eliminate AI ‘hallucinations’. I believe that Apple may be early, rather than late, to the AI game. Expecting that great things are coming from them…

Godwin Josh

Co-Founder of Altrosyn and DIrector at CDTECH | Inventor | Manufacturer

2mo

You talked about the strategic advantage Apple holds with Siri's extensive usage data. Leveraging such rich datasets indeed offers a competitive edge in training LLMs. However, the challenge lies in ensuring data privacy and compliance with legal regulations while extracting insights from these vast repositories. Considering this, how do you envision balancing innovation with ethical data usage practices to harness the full potential of AI technologies like Siri?

Like
Reply

To view or add a comment, sign in

Explore topics