Menlo Ventures合伙人Tim Tully表示，这种权力转移在一定程度上要归功于Claude 3.5的进步，以及大多数公司正在使用三个或更多的大型人工智能模型。This year, corporate spending on generative artificial intelligence (AI) soared 500% to $13.8 billion.
The Zhitong Finance App learned that according to data released by venture capital firm Menlo Ventures on Wednesday, corporate spending on generative artificial intelligence (AI) soared 500% this year, from 2.3 billion US dollars in 2023 to 13.8 billion US dollars. The report also found that OpenAI's market share in the field of enterprise artificial intelligence fell from 50% to 34%. Anthropic's market share doubled, from 12% to 24%. According to the report, this result comes from a survey of 600 IT decision makers in companies with more than 50 employees.
Menlo Ventures is an Anthropic investor.
Menlo Ventures partner Tim Tully said this shift of power is partly due to the advancements of Claude 3.5 and the fact that most companies are using three or more large-scale artificial intelligence models.
Tully explains, “The developers are very savvy — they know how to quickly switch between models. They choose the model that best fits their use case... most likely Claude 3.5.”
Meta Platforms (META.US)'s market share remained at 16%, and Cohere's market share remained at 3%. Google's (GOOGL.US) share rose from 7% to 12%, while Mistral's share dropped one percentage point to 5% in 2024.
The report found that basic models — such as OpenAI's ChatGPT, Google's Gemini, Anthropic's Claude, etc. — still dominate corporate spending, and large-scale language models received 6.5 billion US dollars of corporate investment.
Menlo's report is optimistic about artificial intelligence agents, which are the cutting edge of artificial intelligence and a hot investment area in 2024. Google, Microsoft (MSFT.US), Amazon (AMZN.US), OpenAI, and Anthropic are all studying this technology. Artificial intelligence agents are viewed as a further technology than chatbots. They can perform multi-step, complex tasks on behalf of users and generate their own to-do lists, so users don't have to walk them through the entire process step by step.
“Proxy technology is real; it's not a hype,” Tully said. I don't think it will cure cancer, but it can improve people's productivity and help the company generate revenue. ”
The report found that code generation is the main use case for generative artificial intelligence, with more than half of respondents citing it as a primary use. This was followed by support for chatbots, at 31%, followed by enterprise search and retrieval, data extraction and transformation, and meeting summarization.
