Google: Greater Integration of AI in Search Will Increase Cost Risk, Says Morgan

On Tuesday, Alphabet (NASDAQ: GOOGLE ), Google’s parent company, unveiled its new search assistant, Bard, a rival to OpenAI’s ChatGPT.

The company has stated that it will begin rolling out Bard in the coming weeks, opening it up for testing with trusted users, before making it available to the general public.

The tool is based on LaMDA, a language model for dialogic applications developed by Google itself.

Reacting to Google’s new AI chatbot, Morgan Stanley (NYSE: MS ) analysts told investors the bank believes the tech giant has the technology and scale to maintain/grow its user base.

However, a tighter-than-expected integration “increases the risk of cost overruns, as we estimate that for every 10% of searches that switch to language models, operating expenses increase by about $1.2 billion.” ”.

“The race for artificial intelligence is on,” the analysts added. “Our work on natural language queries suggests they can be up to five times more expensive (on average).”

The bank’s analysts explained that the computational intensity of natural language models that store, retrieve, analyze and compile large amounts of text in the form of responses in a natural language format is very high.

The bank’s updated analysis of ChatGPT’s model size, its computation time, average words generated per query, Nvidia’s Azure A100 GPU (NASDAQ: NVDA ) pricing tiers, and an estimated 50% Azure gross margin (for this analysis), leads them to believe that the average extra GOOGLE natural language cost per query is likely to range from $0.0022 to $0.0220.

“Early demonstrations show how the computational costs of natural language are likely to be higher. But, in our view, the greatest risk with this technology comes from the greater potential for additional costs due to deeper-than-expected integration of natural language into basic search results,” the analysts wrote.

They concluded that if 50% of queries are integrated with natural language in 2024, the additional costs would add up to 6 billion dollars.

Reference source: investing.com