ChatGPT is here to change the world, and it is doing a good job. But there is a small window of exclusivity before competitors appear on the arena. In addition, the Chinese company Baidu has announced that its own version of ChatGPT, with better performance, will be launched this month. This will be the first competitor to ChatGPT, but more are on the way. As you can imagine, Google can’t wait. It will soon launch a similar product. In this context, we must remember Google’s recently launched intelligent chat robot called Bard. Unfortunately, Bard’s performance is not yet satisfactory. It also costs the company a lot of money to run.
Not long ago, John Hennessy, chairman of Alphabet, Google’s parent company, said in an interview that the cost of an artificial intelligence chatbot for large language models could be more than 10 times that of traditional search engines. So they should work hard and do their best to reduce the cost of such operations.
According to Morgan Stanley, Google had a total of 3.3 trillion searches last year. The average cost per search was 0.2 cents. If an artificial intelligence like ChatGPT can handle half of the queries it receives with 50-word answers, Google’s cost will be 10 cents. A $6 billion increase is likely by 2024.
The main reason why artificial intelligence is so expensive is that it requires more computing power. AI relies on chips that cost billions of dollars. In addition, power consumption will also lead to higher costs. Finally, it will also put pressure on the company’s carbon emissions indicators.
Some experts believe that an effective way to reduce costs is to use smaller artificial intelligence models and apply them to simpler tasks.
OpenAI’s computer scientists have found a way to optimize the cost of reasoning through code, improving the chip’s operational efficiency.
In addition, OpenAI announced the launch of a paid version of ChatGPT Plus to generate revenue. It will cost $20 per month. Features of the paid version include skipping queues during peak hours, fast response times, and priority access to new features.
