This audio is automatically generated. Please let us know if you have any comments.
Businesses are optimistic about the potential of generative AI, with plans to invest more resources in initiatives and embark on enterprise-wide changes around its adoption.
less than two years after the launch of OpenAI’s ChatGPTgenerative AI is already the most common use of artificial intelligence, according to Gartner research.
“We really need to help our teams, our business partners, our executives, who get swept up in the hype, to know where this is really helpful and where other things could be better,” he said. Rita Sallam, Distinguished VP Analyst and Gartner Fellow on the Data and Analytics teamat the Company IT Symposium/XPor last week “We really have to keep the organization on the ground to know when it makes sense.”
CIOs must clearly communicate and explain when technology is an effective solution and when it would be better to try other options, such as knowledge graphs or reinforcement learning. After all, organizations rely on the expertise of technology leaders avoid costly missteps.
Failed technology projects can damage an organization’s reputation, customer relationships, and the bottom line. Organizations deploying AI in 2023 spent between $300,000 and $2.9 million in the proof-of-concept phase, and many generative AI experiments never make it past the nascent stage, according to Gartner research.
Sallam said generative is generally not the best tool for companies to:
- Plan and optimize
- Predict and forecast
- Make critical decisions
- Run autonomous systems
Businesses are full of potential use cases, but choosing the ones that bring the most value and the least risk is key.
Weaknesses in generative AI, including unreliability, a tendency to hallucinate and limited reasoning, can derail many use case ideas, Sallam said. Technology leaders can turn to other forms of artificial intelligence, such as predictive machine learning, rule-based systems, and other optimization techniques, to achieve better results.
According to Sallam, large language models struggle to perform accurate calculations, making it difficult to use generative AI for use cases like marketing allocation or route optimization. Instead, CIOs can use knowledge graphs and composite AI, defined as a combination of AI techniques. The guardrails needed to ensure responsible and safe use of the technology can hinder experiments like automated trading and agents. Reinforcement learning would be a better route, Sallam said.
Wrong place, wrong task
Generative AI thrives on content generation, knowledge discovery, and conversational user interfaces. This has spurred countless solutions targeting text and coding, question and answer systems, knowledge management, and virtual assistants.
Companies have been seduced. only 6% of the organizations have Deferred generative AI investments, according to capgemini survey published in july.
“Don’t get me wrong, I think the potential is huge” salam he said But the hype has pushed leaders to over-focus, and potentially over-invest, in generative AI at the expense of the business, according to Sallam.
“Hype is dangerous,” Sallam said. “Organizations that focus solely on generative AI can risk failure in their AI projects and miss out on many important opportunities, so we want to make sure that the hype around generative AI doesn’t take away the ‘room oxygen’.
Recently, vendors have emphasized AI-powered agents with autonomous capabilities, for example. False i know has announced agent capabilities in existing solutions in recent weeks. Salesforce moved its Agentforce platform in general availability this week. Microsoft plans to add agents Copilot Studio next month.
“Now we hardly hear them talking about co-pilots,” Sallam said. “They’ve passed on to agents, and that’s a promise … but the reality now is that it’s still a work in progress. You still have to be careful.”
CIOs must consider how autonomous capabilities fit governance and risk management frameworksespecially as companies emphasize the importance of control and human intervention. Sallam said techniques such as reinforcement learning offer an alternative for powering autonomous systems.
Technology leaders should also exercise caution in use cases that could introduce bias-based risk. Rule-based systems and composite AI offer a more reliable option, Sallam said. Incorporating generative AI into critical hiring or loan allocation decisions could create a recipe for disaster.
“You don’t want to leave that to your big language model,” Sallam said.
