According to Domino Data Lab’s survey from the REV 4 conference, 90% of data scientists think generative AI hype is justified. Respondents are professionals who are leading, developing and operating generative AI initiatives across Fortune 500 companies.

“The findings validate the incredible business potential of Generative AI and its expected near-term impact,” wrote Kjell Carlsson, Domino Data Lab’s head of data science strategy & evangelism, in a blog post. “However, it also confirms key challenges — governance, control, privacy and fairness — as well as the severe limitations of the current, commercially available Generative AI offerings.”

The San Francisco-based Domino Data Lab collected responses from 162 data science executives, data science team leaders, data science practitioners and IT platform owners. Some additional opinions in the report were sourced from Domino Data Lab customers.

55% of data science professionals think AI will have a significant impact on business

More than half (55%) of the data science professionals and IT platform owners surveyed said generative AI will have a significant impact on their business within the next one to two years. Additionally, almost half of the respondents (45%) believe the hype is only rising, expecting generative AI to have an even greater impact than today’s expectations suggest.

Other data from G2, EY and others show the same large impact of AI. In a recent survey of tech executives, CNBC found that AI is their top priority for tech spending over the next year, starting in June 2023; the second priority is cloud computing.

According to Statista, artificial intelligence startups (a category in which Statista includes machine learning, robotics, neural networks and language processing) received a total yearly investment of $5 billion from 2020 to 2022.

Most data science leaders prefer to modify third-party AI

Most (55%) of the data science professionals and IT platform owners Domino Data Lab surveyed prefer to use foundation models from large third parties like OpenAI, Microsoft or Google but create different experiences for their customers on top of the base model. Another 39% want to build their own proprietary generative AI from scratch. Just 6% want to use AI features solely planned and provided by independent software vendors and other third parties.

The respondents believe the biggest problems with commercially available generative AI, such as ChatGPT, are security (54%), reliability (44%) and IP protection (42%).

These concerns mean that organizations need to invest in tools to make it easier to fine-tune generative AI models, as 41% of those surveyed plan to do. Some (35%) also plan to implement governance capabilities for tracking and managing the development of those AI models.

Governance and bias are the top barriers to AI adoption

Data leakage is another problem cited by survey participants. Some are concerned about generative AI having low accuracy or leading to bad business decisions (35%) and budget overreach (33%).

Senior leadership in particular cited concerns about governing generative AI solutions generally (76%), as well as the reliability (76%) and security (71%) of solutions on the market today.

Is generative AI approaching the peak of its hype?

Other industry experts are warning the tech world to temper the hype.

“AI has great potential, but it is a huge high-risk bet, and a large percentage of your investment will likely go nowhere,” said Saurajit Kanungo, president of the consulting firm CG Infinity, in an email. “Only invest if you can measure the ROI in business terms – is it going to decrease costs or increase revenue?”

He points toward Gartner’s 2022 AI Hype Cycle graph, in which generative AI approaches the point labeled Peak of Inflated Expectations.

“I absolutely believe that AI (including generative AI) has the potential to drive value for every organization, big or small. However … I would advise executives to adopt AI as an evolution, not a revolution,” Kanungo said.

He finds the case for generative AI to be stronger than the case for the last hot technology investment trend: cryptocurrencies. “Cryptocurrencies require a whole new ecosystem or market to be made. Business cases to justify investing in generative AI in an organization are an easier challenge compared to making a whole market with cryptocurrencies,” Kanungo said.

Curled from TechRepublic