Forrester: Finding AI talent is challenging

A new report from Forrester on the rapid adoption and evolution of artificial intelligence reveals that organizations are being challenged to hire workers with the skills needed for AI-specific roles.

While the role of the traditional developer will certainly change – and perhaps even be de-prioritized – the report found three areas in which skills will be prioritized. Those are AI developers and engineers, cloud-related roles and data-specific roles. 

By and large, companies are confident that they can find the AI-specific skills they need, as well as find and retain workers skilled in cloud computing and systems administration, according to the report. This is ironic, as for years organizations have been saying they can’t find enough specific software development skills to help them achieve their business goals, yet they’re looking for more skills that are hard to find as of now.

Fiona Mark, a principal analyst at Forrester and primary author of the report, told SD Times that they are seeing a shift away from traditional software development engineers as organizations are redirecting their investments into funding AI. 

“We probably had artificial demand for software development at the peak of the pandemic, so we’re seeing retrenchment from that,” Mark said. “We’ve also seen layoffs in almost every organization across software development. But we also see that there are skills within the traditional software development space that are being transferred over into the AI space. So it’s not like we don’t need coders anymore. What we need is people who can work with AI, who’ve either upskilled themselves or gained experience through higher education or through previous roles in other organizations.”

Mark made it clear that when we think about an AI engineer, it’s not necessarily someone who is doing prompting – asking GenAI, for example, to write some code based on requirements. It’s about finding talented people who can put in place the right training data to train AI, and ensuring that the AI is safely and responsibly acting within organizational controls. That’s where organizations, she said, are “wanting to take some of those [large language models] and make them specific for their needs. It’s where you think about, ‘how do I train this AI to behave in a way that’s right for my organization, to support my employees or to support my customers.’ “

It’s also important that these AI engineers can work with things like TensorFlow or PyTorch, and other technologies that are being used to build AI models and to implement them, she said.

Dealing with more and more complexity

One of the benefits of AI that is being touted is its ability to help developers deal with the complexity of software today, and that also dovetails with the changing roles of cloud architects and data architects. Those individuals and teams will be tasked with architecting and building systems to reduce complexity and take advantage of the investments in AI, while enabling developers to shift away from managing complexity to creating business value.

From a software developer standpoint, copilots and other tools are helping reduce some of the overhead that developers are facing, and Mark said developers and software engineers will be spending more time thinking about those complexity questions around the architecture as to how effective a solution is. “It’s more than just, does the solution work,” Mark said. Developers, she added, “should be getting out of some of the more repetitive elements of their developer role.”

Meanwhile, organizations requiring these new skills have less of an appetite for bringing on people early in their careers. The idea is that AI will be able to pick up what those early career people – those with less than four years’ experience – know. “And when that starts to happen, we have a pipeline problem and a management problem coming down the line, you know, in four to eight years when we don’t have some of that talent moving up the organization, and we don’t have access to some of the resources that we need.”

Higher learning to gain skills

It has been said for years that colleges and universities are not training developers in specific programming languages, software architecture and more. But with AI experimentation and adoption exploding, colleges are looking at artificial intelligence not only just from a computer science standpoint, but also from a mathematical point of view. And while boot camps have always helped upskill developers, Mark said they no longer are able to offer the same level of job placement that they were two or three years ago. 

“What we’re seeing is the market is very much becoming quite split in that people who have AI skills are very much in demand, and people who are more generalist are really struggling to find work,” Mark said. “And the other piece we still see, from a tech role point of view, we’re not seeing the four-year degree as a formal barrier to entry for many roles. We’ve seen that sort of requirement being removed. But that doesn’t mean that it’s not an implicit requirement, given the nature of AI and these specialist skills that we’re now seeing in the marketplace.”

Cloud native and data skills

AI demands a lot of computing power, and organizations are looking at how they can manage and deploy the right resources in a way that is cost effective and that allows for taking advantage of these new technologies. So cloud architecture and cloud engineer roles will see greater demand, Mark said. 

“Cloud is very mature now, but for many organizations, they’re still in the process of, we migrated to cloud platform services a decade ago, but we did it in a kind of lift and shift. Now we need to re-architect in terms of taking advantage of what technologies are available within the cloud, what capabilities there are,” she explained. “This infrastructure is where we’re going to be for the foreseeable future.”

Data management also becomes a huge issue in an AI world. All AI models are using data to learn from, to generate, and to predict. “This is what you build everything on,” Mark said. “And so having that data in a way that is well managed, that has high quality integrity, that is well governed, with appropriate security and privacy controls, all of those elements are foundational to being able to take advantage of your AI.”

The post Forrester: Finding AI talent is challenging appeared first on SD Times.

from SD Times https://ift.tt/CwbiNtJ