Digital Platform Policy Highlights - Digest 36
Q2 2024 to Q3 2024 policy changes: This post highlights how major platforms are evolving their strategies by increasing royalties for creators and selling data access for AI training.
This post is part eleven of a series documenting policy changes and feature improvements introduced by platforms in Q1 2024 to Q3 2024.
TL; DR→ Here are policy changes to utilize increased computational capabilities as well as advances in AI:
Apple Music’s attempt at winning using Spatial Audio
Apple Music will offer artists a bonus payout of up to 10% in royalties for songs available in spatial audio. This policy change incentivizes musicians to produce content in Dolby Atmos, enhancing the immersive listening experience for users. If artists warm up to producing Spatial Audio, Apple is uniquely placed to differentiate its music service from competitors like Spotify. Leveraging Apple’s hardware ecosystem, i.e., AirPods, to deliver exclusive audio experiences may convince some owners of AirPods to consider Apple Music even if they otherwise strictly prefer Spotify’s offerings. (link)
Reddit sells data access to OpenAI
Reddit’s deal with OpenAI grants the AI company real-time access to Reddit posts to train its ChatGPT model. Reddit has also made a similar deal with Google. While this partnership is a win-win for OpenAI (which receives training data) and Reddit (which presumably receives money), Reddit’s posters will not be compensated or be asked for consent in having their data used. This rush to monetize user data through AI training comes at a critical moment - as models become increasingly capable of learning from everyday conversations, not just expert knowledge, platforms are racing to secure data access before potential regulations on data ownership and user compensation take effect. (link)
OpenAI and the “App Store” for custom GPTs
A confusingly named GPT Store allows developers to build customized ChatGPT versions for subscribers of the ChatGPT Plus service.. This represents a strategic move to abstract away underlying AI complexities - users can access specialized tools for any generative task (writing, coding, image creation, talking in parseltongue) through a single interface, rather than juggling multiple AI services. By positioning ChatGPT as platform for “everything generative”, OpenAI is attempting to abstract out what smartphone operating system you are using, similar to how operating systems abstracted out which hardware you used. (link)
OpenAI’s SearchGPT as a Google Competitor
Now available to everyone logged in, the AI-powered search engine was initially available through a waitlist, with no immediate plans for monetization. Similar products from Google (Gemini and Search Generative Experience) have come under criticism for providing false or misleading information, but as a new product, SearchGPT may have more leeway when it comes to accuracy. This marks a significant shift in how users might access information - while traditional search engines present data for users to evaluate, GPTs act as great persuaders that invariably hallucinate, potentially blurring the line between information discovery and AI-guided interpretation. (link)
Google deploying Generative AI in Cloud Security Tools
Cloud giant has added Gemini AI capabilities to its security suite, aiming to enhance threat detection and automate security operations, after having bought the cloud security firm Mandiant for $5.4 billion last year. While Google's solution relies entirely on cloud processing, it's likely that tech giants will eventually offer hybrid solutions using locally-deployed GenAI to process sensitive data on-premises. Processing sensitive log data on-premises before sharing only sanitized insights to the cloud is a crucial capability for organizations with strict data sovereignty requirements. (link)
Snapchat Launches New AR and ML Tools for Brands and Advertisers
Snapchat leverages the increased computational power of modern smartphones to deliver complex Augmented Reality (AR) experiences like virtual-try ons. These tools use on-device machine learning technology for ad targeting and personalization. This move aims to strengthen Snapchat's position in the digital advertising space by offering brands unique ways to engage users, while making use of the increased capabilities of smartphones to deliver the experience. At least the early adopters of glasses would be thrilled about such personalized content (link)
Research help from Marshall Singer, Anantesh Mohapatra, Jennifer Xie, Anna Li and John Mai (Thanks a ton, folks!)
Thank you for reading Platform Policy Research. If you know someone who like this stuff, please share it with them :)
Super relevant summary!