Last year, Meta predicted it would earn $2-$3 billion in revenue in 2025 and $3 billion in revenue by 2035. According to court documents that were not sealed on Wednesday,.
The documents submitted by the authors of the book where lawyers suing Meta are claiming it is fraudulent training for the company’s AI on their work, and do not show exactly what Meta considers as a “generative AI product.” However, it is common knowledge that the tech giants will stand to make money from AI generated with many flavors and make more money.
META has earnings sharing agreements with certain companies that host model collections for Open Llama Collections. The company recently launched an API to customize and evaluate llama models. The company’s AI assistant, Meta AI, may ultimately display ads and offer subscription options with additional features, CEO Mark Zuckerberg I said Calls will be made Wednesday during the company’s first quarter revenue.
Court documents also reveal that Meta is spending a huge amount on AI product groups.
In 2024, the company’s “genai” budget exceeded $900 million, potentially exceeding $1 billion this year. This does not include the infrastructure required to run and train AI models. Meta previously said it plans to spend between $600 billion and $80 billion in 2025 primarily on vast new data centers.
These budgets could have been high if they included transactions licensing books by authors suing meta. For example, Meta was discussed in 2023 over $200 million to obtain training data for Llama. About $100 million of that would have been headed exclusively for books according to the documents. However, the company is said to have decided to pursue other options: pirated large e-books.
Meta did not immediately respond to requests for comment.
TechCrunch Events
Berkeley, California
|
June 5th
Book now
