AI won’t replace developers—but it will change who gets hired
AI is transforming coding from a technical task into a multidisciplinary challenge.
As technology advances and artificial intelligence (AI) takes over more tasks, human skills—often referred to as ‘soft skills’—are becoming increasingly important in the workforce. In the past, mastering programming languages was enough to secure a development position at a tech company. Today, however, the reality has shifted. The requirements now extend beyond technical knowledge and familiarity with AI tools; a broad worldview, the ability to understand and analyze human behavior, management and teamwork skills, and the capacity to quickly learn emerging tools and technologies are essential. We spoke with industry executives to understand which skills are truly critical for developers in this new era.
"Development is not just about knowing code—it’s about understanding user experience, data, and human psychology," says Uzi Yaari, VP of the Digital Division at Elad Software. According to Yaari, today's programmers must think about how users would want to act, search, and choose because, in the end, AI aims to simulate human thinking. "Anyone who thinks that development is just about writing code is mistaken. The best programmers are those who know how to think like the user. Developers are typically great at talking to machines but not always with people. Today, because of AI, there is a psychological and behavioral dimension to consider," he explains.
Beyond technological expertise, developers today must understand how humans make decisions and incorporate these insights into their models. "We test and practice this through various simulations, and it’s one of the most critical skills in working with AI," says Yaari.
He views the skills required of developers today as part of a three-tier model: at the base is coding and technical knowledge, followed by data science, and at the top, machine learning and algorithms. While AI-generated code will eventually become more prevalent, it’s not yet at the point where developers can rely solely on it. As such, in-depth mathematical knowledge remains critical. "We are only at the beginning of the AI journey. We are working a lot on development and testing with AI tools, and the results are promising, but AI-generated code is still not good enough. Therefore, human abilities must be strong," he says.
AI Managers Needed
Software developers today are expected to master basic AI tools, like ChatGPT and CoPilot, as part of their work process. However, more advanced tools that can generate entire code segments and automate development processes require a different level of skills—primarily management skills. “To work effectively, you need to explain the architectural design of the product to the AI system at a higher level. It’s like having four junior developers working for you; you need to explain the product to them to get the best results," says Amir Ish-Shalom, VP of R&D at Lightrun. At Lightrun, they ensure that experienced team leaders or architects are the ones using AI tools to define how they should be used correctly and effectively.
According to Ish-Shalom, working with AI requires not only technical knowledge but also a deep understanding of the available tools and the ability to integrate them effectively into development processes. "Advanced tools like Cursor or Windsurf are essentially AI agents that can generate entire parts automatically. While these tools are powerful, they require the ability to understand and control the results to avoid inefficient or incorrect code," he says.
One challenge with advanced AI tools is that junior developers sometimes rely too heavily on them without fully understanding the code being created, which can lead to bugs and compromise product quality. "We’ve observed that while the speed of production is higher, more problems are created, ultimately delaying progress. You need the ability to solve bugs and understand the product built by AI-generated code. We often see people who have completed projects that look impressive, only to realize they have no idea how the AI-generated code works," says Ish-Shalom.
A Different Mindset
“We’re looking for candidates who know how to leverage AI not just to write code, but to create real business value,” explains Amram Guttman, SVP of Development at Playtika. According to Guttman, the major shift from a decade ago is that technical ability is no longer enough. "In the past, recruitment for development positions focused mainly on programming languages, previous projects, and years of experience. Today, the picture is more complex. Along with a solid technical background and the ability to explain technological choices made in past projects, we place significant emphasis on in-depth knowledge of AI tools and understanding the business value they bring."
"We’re not only evaluating the ability to use AI tools like automated code generation, testing, and code reviews, but also developers’ ability to explain their choices, understand the limitations and benefits of these tools, and use them effectively. A key challenge is the overreliance on AI tools without understanding the broader business and technological context, which can lead to errors. From our perspective, AI is a tool that supports, but doesn’t replace, a solid understanding of programming fundamentals and business thinking," he says.
In fact, beyond technical AI skills, startups today require an AI mindset—an approach focused on identifying opportunities for systemic improvements through AI. "It’s not about knowing how to use models; it’s about integrating them into work processes and leveraging them for the team, organization, and product. In other words, we’re looking for 'builders'—developers who think about the entire system, not just the code they write," says Dror Grof, partner at Team8 and leader of its Ideation team.
Grof adds that when the right people are found, soft skills become just as important as technical abilities, particularly two key skills: mentoring and resilience. "We seek people who naturally want to teach and develop others, those who can build tools to help others improve. This is the difference between giving fish and teaching how to fish. The second is resilience—developers who embrace change and see it as an opportunity to grow and explore new areas."
There is consensus among recruiters about the importance of adaptability, professional flexibility, curiosity, and a willingness to learn. "When we recruit experienced developers, we expect them to already be using AI tools, to be able to explain their technological choices, and to understand the impact of these tools on performance and efficiency. For junior developers, we look for curiosity, a willingness to learn, and the ability to write code independently without relying solely on AI. We also value mental flexibility and openness to continuous learning in this rapidly changing environment," says Guttman.
The Changing Role of Developers
To understand the demands on development teams in the age of AI, imagine a pyramid of skills: at the base, mastery of code and programming; above it, in-depth knowledge of data science; and at the top, specialization in algorithms. "Mastery of algorithms provides developers with the infrastructure necessary for AI development in a machine learning environment," says Yaari. "In the end, developers are given a project specification but must know how to turn it into calculations and practical work on their own."
However, as new technologies emerge daily, the ability to work in a dynamic team has become just as critical as technical knowledge. "It’s no longer a developer sitting alone in a niche, getting a project brief, and working in isolation. Teamwork is essential, and that’s one of the things we look for in job candidates," says Yaari. Additionally, the ability to think conceptually and abstractly has become critical. "You need to envision how things will evolve throughout the project," he adds.
Technological skills, such as working with algorithms that include deep learning principles, natural language processing, and computer vision, are also required at companies like Agmatix, an ICL startup specializing in digital solutions for agriculture. "Problem-solving skills are essential, including identifying problems, formulating solutions, and selecting the appropriate algorithms and tools to solve them. Critical thinking—questioning how AI models were developed, what data was used, and their potential impact—is also crucial," says Liron Prizant, VP of HR at Agmatix.
AI-Native Companies
The rise of AI-native companies is transforming the skill set required of employees. "AI-native companies are built with AI at their core. It’s not an add-on but a fundamental part of the company, culture, product, and business model," says Grof. These companies need employees with a systems mindset—those who not only know how to use AI technologies but also identify areas for optimization and performance improvement. They seek employees comfortable with uncertainty, who understand that change is inevitable and who believe in experimentation, rapid testing, and fast learning.
Lastly, Grof notes that employees must be able to work with AI tools beyond their narrow areas of expertise. In startups, employees who were once limited to specific skill sets will need to embrace broader roles supported by AI capabilities.
