To The Moon

Google Search is much cheaper than the AI-based search

Alphabet's Google has been facing a new challenge in its pursuit of generative artificial intelligence (AI) that can power chatbots like OpenAI's ChatGPT: the cost. As the computing power required for ChatGPT-like AI is enormous, it puts a significant financial burden on the company. The computing power involved in AI, such as ChatGPT, is much higher than conventional search, with billions of dollars of chips required for inference.

Analysts saying that even if companies like Google and Microsoft managed to extract revenue fromchat-based search ads, the technology could chip into Alphabet's bottom line with several billion dollars of extra costs.

Alphabet's Chairman John Hennessy also confirmed that the large language model AI will comst the company ten times more than a standard keyword search. Morgan Stanley claims that Google's 3.3 trillion search queries last year cost roughly a fifth of a cent each, a number that would increase depending on how much text AI must generate.

Why so expensive?

This form of AI is pricier than conventional search is the computing power involved. That computing can only be achieved by billions of dollars worth of chipsets. Electricity also adds to the costs.

The high cost ofChatGPT-like AIcan turn out to be a challenge for Alphabet's Google, which is under pressure to develop AI-powered chatbots that can rival Microsoft's Bing search engine. The company will try to optimize the costs before making Bard widely available. Some analysts also believe that with increase in scale, the costing should become more affordable for companies.