Meta has its own AI news, including release of open source text-to-speech and speech-to-text models support more than 1,100 languages. But mostly they’ve trying get their business in order. They took another step in that direction yesterday by selling GIPHY, which they bought for $400 million in 2020 but couldn’t keep due to anti-trust issues. Buyer was Shutterstock, which paid $53 million.
Microsoft kicked off its Build conference for developers yesterday. The focus wasn’t on marketing but there was still a lot of relevant news, including use of Bing as the default search engine for ChatGPT, chat interface-driven tools for website building and app creation, and still more chat-driven data preparation, cleaning, and analysis. It seems the next step beyond low-code development is voice-based coding, or vo-code.
It’s barely one year since Meta closed its $1 billion acquisition of the Kustomer CRM platform, which had been delayed since 2020 due to regulatory reviews. But Meta’s fortunes have changed since then and, like a family that bought a boat they couldn’t afford, they’ve now sold it back to the original owners for a fraction of what they paid. They did keep a minority stake.
GrowthLoop will use TransUnion identity data to improve U.S. consumer match rates on client files sent to advertising media such as Facebook and Google Ads. The data should also help clients to find matches among records within their files, building more accurate customer profiles. GrowthLoop calls itself a “composable CDP”, meaning it works with data assembled in external data warehouses.