Apple Cuts Deeper Than Facebook into Augmented Reality Platform Market

The article discusses how Meta, the parent company of Facebook and Instagram, is facing a challenge from OpenAI’s GPT-4 language model. According to court filings, Meta executives were obsessed with beating GPT-4 internally, which suggests that they were trying to create their own AI language model that could rival OpenAI’s.

However, the article also notes that Meta’s own language model, Llama, is still in its early stages and has been criticized for its limitations. The article mentions that Llama’s podcast hosts had to be taught not to act annoyed at humans, indicating that the model still needs a lot of work before it can compete with GPT-4.

Additionally, the article discusses how TikTok users are trying to migrate to Chinese app RedNote, but it seems that this effort is not going well. The app has been plagued by issues such as poor user interface and lack of features.

The article also mentions that Google’s NotebookLM AI had to be taught not to act annoyed at humans when interacting with them, similar to Llama. This highlights the challenges that companies are facing in creating AI models that can interact with humans effectively.

Finally, the article notes that Meta plans to cut roughly 5% of its workforce based on performance, which suggests that the company is trying to streamline its operations and focus on more productive areas.

Overall, the article provides an update on the latest developments in the tech industry, including the challenges faced by companies like Meta and OpenAI in creating effective AI models.