Building generative AI applications is too hard, developers say

“According to respondents, productivity (42%), flexibility (41%), ease of use (40%) and integration (36%) are the four most important qualities of enterprise AI development tools,” Gunnar noted. “However, more than a third of those surveyed also said those same traits were the rarest.”

About a third of respondents also deplored four other important characteristics: quality of documentation, cost-effectiveness, community support and resources, and the tools being open source.

Considering the number of tools needed to get the job done, it’s no surprise that developers don’t want to spend much time adding yet another one to their arsenal. Two-thirds of them are willing to spend just two hours or less learning a new AI development tool, another 22% devote three to five hours, and only 11% devote more than five hours to this task. And in general, they are not inclined to explore new tools very often – only 21% said they check new tools monthly, 78% do so every six months to a year, and the remaining 2% rarely or never. The survey found that they tend to view about six new tools each time.