NVIDIA has seen their valuation 5x over the last 5 years and is now worth more than $2 trillion. In the latest episode, we dive a few layers deep into the company. We start with the origin of the company and the launch of their first GPU in 1999 which unlocked several computing-heavy applications including gaming and video editing. We then talk about the NVIDIA of today that generates 78% of revenue from their AI-focused data center business (compared to 40% in 2021). Finally, we analyse what assumptions need to be true for the company's high valuation to hold up including long-term AI demand and competitive landscape.

Here are the highlights from the podcast episode. You can find a longer form analysis on both topics in the podcast episode - Spotify | Apple Podcasts. The podcast follows the same core principles as the writing (primarily analysis driven with occasional opinions sprinkled in). Just reply to the email if you have any feedback or topic suggestions.

Last week, 20 tech companies across Big Tech, social networks and AI providers signed a voluntary pledge to fight election interference from AI-generated content going through 2024. We dive into the timing of the announcement (2024 is a big election year across the globe), some recent deepfake mishaps, where should liability lie when bad things happen, and whether this pledge is enough. Bonus - couple of ideas for AI startups.

Last week, Meta announced that they will not proactively recommend political content across their recommendation surfaces including Reels and Threads. We dive into criticism Meta has gotten in the past, what led Meta to this decision, the increasingly dispersed media landscape today, and how this decision impacts political discourse especially in an election year.