The ChatGPT operator, which made $28 million in revenue last year, could be reaching the $100 million revenue mark each month in 2023.
Since partnering with Microsoft, OpenAI’s GPT large language models have made it not just an integral piece of Microsoft’s latest products but the de facto LLM for many companies interested in exploring generative AI.
Nod.ai will let companies run AI models specifically tuned to AMD’s chips.
AMD has been trying to increase its market share against Nvidia, currently the de facto leader in AI chips. Most large language models run on Nvidia’s H100 GPUs which are now in high demand.
AMD CEO Lisa Su spoke with The Verge’s Nilay Patel during the Code Conference and said the company sees a lot of opportunity in providing chips that can run AI models to smaller companies that may not have direct access to Nvidia’s GPUs.
Most governments are still in the planning stages of figuring out how to regulate generative AI. This includes Australia, Great Britain, the United States, and the United Nations, and the greater G7. The European Union, listed in the Reuters report as planning regulations, is discussing draft rules.
Countries like France, Italy, Japan, Poland, and Spain are investigating complaints about data breaches around OpenAI’s ChatGPT.
Only China has implemented interim guidelines around generative AI.
Amid the shortage driving prices of Nvidia’s H100 chips up, Reuters reports ChatGPT maker OpenAI may build its own chips, work closer with Nvidia, or even explore other sources. Making its own chips is a risky play that could take years to pay off unless it buys a smaller player.
Everyone wants to shake Nvidia’s dominance in the space, with Microsoft, Amazon, and Google developing chip technology of their own. Meta is also in the early stages of making its own chips.
