Successful deployment of enterprise language models requires overcoming key challenges to training them effectively, responsibly and securely.
boston, May 11, 2023 /PRNewswire/ — Now that the hype around Open AI’s ChatGPT has led 45% of executives to increase investment in artificial intelligence (AI)*, new Expert.ai launches large-scale language models: Opportunity, Studying risks, and paths. Forward revealed that more than a third (37.1%) of companies already plan to train and customize language models to meet their business needs.
The majority of companies (78.5%) recognize that the effort required to effectively train usable and accurate company-specific language models is a significant undertaking requiring dedicated resources and budgets doing. Nearly three-quarters of the companies surveyed have, or are considering adding to, the budget to support large-scale language model (LLM) deployments.
“Enterprise-specific language models with a human-centric approach are part of the future.” To tell Marco Varone, Founder and CTO of Expert.ai. “Business natural language use cases always require some degree of domain-specific training applied to existing proprietary or open source LLMs. It is efficient, fast and resource intensive, with subject matter experts monitoring and reconciling data and inputs throughout the process to ensure accuracy, transparency and accountability.”
The survey found that while only a minority (21.2%) of those surveyed supported a moratorium on LLM training, the majority (70.6%) of AI professionals and practitioners It points out the necessity of using Key deployment challenges included data privacy and security (73.1%), accuracy and quality of deploying production models (51.2%), and knowledgeable resources on how to build and train LLMs (40.7%). increase.
For companies prioritizing AI transparency and accountability, generative AI and LLM can pose real risks to their ESG goals and performance, with truth (69.8%), bias (67.3%), sensitive data leakage (62.6%) is the top concern. .
Regardless of which direction an organization chooses, fundamental AI data governance principles apply to generative AI and LLM. 24.0% of survey respondents said more restrictions should be put in place to test LLMs and clearly communicate applicable policies, while 38.8% said some degree of freedom should be more encouraged 34.3% said the current principles are sufficient.
Download the Expert.ai e-book Large Language Models: Opportunity, Risk, and a Path Forward.
* Gartner Poll Finds 45% of Executives Say ChatGPT Drives Increased AI Investments – May 3, 2023Link
About research
Expert.ai’s “Large Language Models: Opportunities, Risks, and Ways Forward” report summarizes the findings from over 300 business, technical and academic natural language AI experts around the world. increase. The interview was conducted online by expert.ai. April 2023 The goal is to explore the potential opportunities and risks associated with generative AI and provide recommendations for future paths companies can take to develop and deploy LLM.
About Expert.ai
Expert.ai (EXAI:IM) is a leading company in AI-based natural language software. Insurance, banking, finance, publishing, media and defense organizations are experts in transforming language into data, analyzing and understanding complex documents, accelerating intelligent process automation and improving decision-making. We are using .ai. Expert.ai’s purpose-built natural language platform combines simple and powerful tools with a proven hybrid AI approach that combines symbolic and machine learning to solve real-world problems and run businesses at speed and scale. strengthen towith office Europe and North America, expert.ai serves global companies such as AXA XL, Zurich Insurance Group, Generali, Associated Press, Bloomberg INDG, BNP Paribas, Rabobank, Gannett and EBSCO. For more information, please visit https://www.expert.ai.
Source Expert.ai
