Partners Group Holding AG executives think the private equity firm is at the forefront of generative AI utilization among alternative investment firms, already using its own version of ChatGPT to make everyday tasks more efficient, with much more in the works.
The model became available to all employees in early June, said Patrik Bless, Zug, Switzerland-based co-head of business applications and chief information security officer at $135 billion Partners Group, in an emailed response to questions.
At the moment, Partners Group executives are experimenting with using the technology for finding attractive subsectors for investment, identifying its portfolio companies' competitors and requests for proposal writing support and report generation, among other applications, Mr. Bless said. "We are on the lookout for more," he added.
Partners Group technology called PRIMERA GPT has just been upgraded to GPT-4, having started with GPT-3.5 technology, Mr. Bless said. PRIMERA GPT is "ring-fenced," meaning all input data remains under firm control, Mr. Bless said.
Ring-fencing ensures "that our information security cannot be compromised as we develop our use cases for the tool," Mr. Bless said.
As the model gradually improves its literacy levels across all aspects of the firm, the "generative AI technology is likely to become increasingly integrated into working life," Mr. Bless said.
But the use of the technology is in its earliest stages. Governments are starting to focus more on how generative AI is being used, making it hard to predict what the regulatory landscape will eventually look like, he said.
"Ultimately, the use of generative AI needs to maintain the 'human in the loop' to ensure it remains purely a support function rather than a decision-maker," Mr. Bless said.