[ad_1]
Every IPO S-1 prospectus includes a section listing all kinds of risk factors that might make its shares not a good buy. Many of them are typical warnings, financial performance warnings, or acts of war and nature. But some are more company specific.
With ServiceTitan, a cloud services startup Public offering documents With the Securities and Exchange Commission on Monday, we may see the birth of a new standard warning: Large language models (LLMs) that help grow a business can also do harm.
ServiceTitan had a 1,150-word risk factor on how its use of AI, specifically generative AI, could negatively impact its business. She warned that LLM can produce “inaccurate” information and engage in “discriminatory” behaviour; Their LLMs may infringe the copyright or intellectual property rights of others; Using LLMs means exposing more data to potential hacks and damage. But again, if you can’t get enough data, you may not be able to continue offering AI products or build new ones. Furthermore, the company warns that its employees or contractors may accidentally share its customers’ private data with external systems which could then lead to security breaches, or use that data to train their models.
ServiceTitan warns that its use of AI may one day conflict with social and ethical issues. Or that future regulations could cost them money.
ServiceTitan also fears that it may not be able to hire the AI experts it needs for its products, and even if it could, that talent would be expensive. ServiceTitan warns that its AI products rely on third-party services that it does not control — it specifically names Microsoft and OpenAI. So, if they become unavailable or have problems, that’s also a risk.
These caveats are interesting because ServiceTitan operates in the nexus of industries expected to be dominated by GenAI, specifically LLM-based AI agents. ServiceTitan offers software for companies in the field service industry — largely small businesses, such as building contractors, HVAC professionals, landscapers, and the like. Its software performs functions such as marketing, customer relationship management, customer support, and accounting. (ServiceTitan declined further comment.)
AI agents are now aggressively moving toward everything related to sales, marketing, and CRM, so much so that the king of the CRM market, Salesforce CEO Marc Benioff, told TechCrunch that his company alone could have a billion AI agents working in client companies. Salesforce in one year. .
ServiceTitan has been offering AI-powered services since at least 2023, called Titan Intelligence. In October, It has also rolled out a suite of AI agents for sales, customer service and call centers.
But LLMs are designed to study words, images, and data and create their own version of things. That is, they were designed to make things up. OpenAI scientists are busy in the race toward general intelligence, that is, emulating humans’ ability to reason better, which means they are working on artificial intelligence that they hope will be more creative.
The reliability of LLM for commercial use is certainly something that will have to be resolved over time as more companies specialize in agents where reliability, not creativity, is essential. But it’s worth noting that those selling AI bots today are actually saying the quiet part out loud in their legal spiel: At this early stage, AI adoption could create as many problems as it solves.
[ad_2]