A recent report by LangChain, a software framework, surveyed over 1,300 professionals to gain insights into the state of AI agents in 2024. The report revealed that 51% of the respondents are already using AI agents in production, with 63% of mid-sized companies having deployed agents and 78% actively planning to integrate them. Surprisingly, the survey also showed that professionals in non-technical companies are also willing to deploy AI agents, with 90% of respondents in these companies having or planning to put agents in production. This is almost equivalent to the percentage in tech companies, which stands at 89%.
The positive sentiment towards AI agents is further supported by Research and Market’s report on the AI agents market, which predicts a significant growth from $5.1 billion in 2024 to $47.1 billion in 2030, with a CAGR of 44.8%.
However, there is still some confusion around the definition of AI agents. While a majority of the respondents in Langchain’s survey use AI agents for research summarization and personal assistance, a significant 35% use them for coding tasks. This raises the question of whether AI agents should be considered as assistants or agents with absolute autonomy.
Google’s claim that 25% of newly written code is AI-generated has also sparked criticism, with some suggesting that it primarily relies on a code-completion engine. Similarly, Stripe’s recent launch of an SDK for AI agents has been questioned by users, who see it as just a fancier way of referring to API and function calls.
At Oracle CloudWorld in 2024, the company announced over 50 AI agents in its Fusion Cloud Application suite. However, Oracle’s executive vice president of applications development, Steve Miranda, clarified that these agents are not completely autonomous and will require human assistance in the early stages.
Similarly, Ketan Karkhanis, CEO of ThoughtSpot, explained that many systems today, such as Microsoft’s Copilot, operate on single-turn Q&A, answering one question at a time. This suggests that the definition of AI agents is still evolving and will continue to do so as technology advances.
