OpenAI, the artificial intelligence research lab behind ChatGPT, is facing multiple lawsuits over the use of copyrighted material to train its AI model. Publishers, including The Center for Investigative Reporting, The New York Times, and the Author's Guild, are seeking compensation for the use of their works [a07eff20]. OpenAI has signed licensing agreements with news organizations to access their content, but the lawsuits allege that the company has used copyrighted material without proper authorization [a07eff20]. In response to the legal action, OpenAI has deleted two large datasets that were used to train its previous AI model, GPT-3. The company has not yet commented on the lawsuits or provided further details about its plans to address the copyright infringement allegations [a07eff20].
Despite these legal challenges, OpenAI has achieved a significant milestone by generating $3.4 billion in revenue from its ChatGPT subscriptions. The majority of the revenue, around $1.9 billion, comes from ChatGPT Plus, which has 7.7 million subscribers paying $20 per month. ChatGPT Enterprise follows closely with $714 million in revenue from 1.2 million subscribers at $50 per month. OpenAI's API services contributed $510 million to its revenue, while ChatGPT Team added $290 million from 80,000 subscribers paying $25 monthly. Microsoft's sales of OpenAI models on its Azure platform contribute about $200 million annually to OpenAI's revenue [b3199b9f]. Despite its revenue success, OpenAI operates at a loss due to high operational costs, including an estimated $700,000 per day for running AI models like ChatGPT. OpenAI remains optimistic about its future and has appointed new leadership to support its expansion and product development efforts. The company aims for sustained growth in the competitive AI market.
In a remarkable surge, ChatGPT has added 50 million weekly users in just two months, bringing its total weekly user count to over 250 million as of October 2, 2024. This growth is attributed to the back-to-school season and represents a doubling of its user base since November 2023, when it had around 200 million users. The user base had previously reached 350 million monthly users in June 2024 [f88c4f99]. OpenAI has raised $6.6 billion in funding, which is expected to support further growth, especially with potential integrations like Apple Intelligence on the horizon [f88c4f99].
In other AI news, several leading AI companies, including Nvidia, Anthropic, and Apple, have reportedly leveraged subtitles from over 173,000 YouTube videos across 48,000 channels to train their models [eab638e6]. This approach allows AI models to understand human language better by learning from the vast amount of data available on YouTube. Additionally, OpenAI has unveiled the GPT-4o mini, a more compact and cost-effective iteration of its GPT-4o model, replacing GPT-3.5 for all Free, Plus, and Team users on ChatGPT. The GPT-4o mini is priced at ¢15 per 1 million input tokens and ¢60 per 1 million output tokens, representing a 60% cost reduction compared to GPT-3.5 Turbo [eab638e6].
Turing, a staffing firm, has been providing data for AI models to major AI companies, including OpenAI. Turing initially pitched its services to OpenAI, offering computer code to improve the reasoning ability of their language model, GPT-3. This collaboration helped OpenAI make a significant leap in performance with the release of ChatGPT. Turing has expanded its business model to offer AI consulting services and specialized data from various industries. The company hires subject-matter experts to create 'input and output pairs' to improve AI models' reasoning and understanding. The goal is to develop AI models that can carry out complex tasks independently. As AI models become more reliable and specialized, they may work together to create an 'agentic' future where AI becomes a helpful tool for businesses and consumers [bb923b63].