The global race in artificial intelligence and deep technologies is increasing the demand for scientific depth, but many researchers complete their training without experience in product development and without the network of connections that graduates of technological units enjoy. Systematic connections between academia and industry could turn this gap into an Israeli advantage.
By Chaim Geron
The global race in artificial intelligence has changed the type of workers that technology companies are looking for. Alongside experienced software engineers, there is a growing demand for researchers who bring advanced knowledge in mathematics, computer science, physics, biology, and engineering — and who know how to translate it into systems that work outside the lab.
The establishment of large research labs at technology companies and the fight for artificial intelligence researchers illustrate the change. Artificial intelligence systems, advanced chips, quantum computing, computational biology, and other deep technologies do not rely solely on writing code. They require scientific understanding, formulating research questions, analyzing complex data, and critically examining results.
This does not mean that regular software engineering has lost its importance. On the contrary: research without engineering does not become a product, and engineering without a scientific basis sometimes has difficulty breaking through the boundaries of what exists. Advanced companies need a combination of both capabilities.
A scientific asset that is not fully realized
Thousands of doctoral students, researchers, and PhD holders work in the fields of science, technology, engineering, and mathematics in Israel. Years of research give them the ability to deal with problems that have no known solution in advance, to read professional literature critically, to design experiments, and to draw conclusions from partial data.
However, the academic track was built primarily to train researchers. It rewards papers, teaching, and scientific achievements, and does not require familiarity with product development processes, working with large-scale software systems, customer needs, or budget and schedule constraints.
When researchers seek to move into industry, they sometimes find that their scientific depth does not automatically translate into the experience that employers are looking for.
On the other hand, graduates of the IDF's technological units enter the job market with practical experience, familiarity with teamwork, and a professional network of contacts. They have already developed systems under real-world conditions, and often gain quick connections to investors, entrepreneurs, and companies.
This advantage does not necessarily stem from greater scientific depth, but from the applied framework built around them.
A researcher who completed a PhD in his thirties may come with rare expertise and high analytical ability, but without experience working in a production environment, in product development, or in implementing models with users.
This is a gap that can be closed. The problem is not the quality of researchers, but the mismatch between their training and the requirements for entering the industry.
Don't tear down the ivory tower—build a bridge to it
The solution is not to turn universities into vocational schools. Basic research has value in itself, and it needs freedom, time, and distance from commercial pressures. Knowledge commercialization companies also play an important role in protecting intellectual property and transferring inventions from academia to industry.
However, additional tracks can be created for researchers interested in integrating into applied development. Such tracks could include training in machine learning and deep learning in a production environment, working with data and cloud infrastructures, internships in research and development teams in companies, and mentoring by industry professionals.
It is also worthwhile to expand scholarships and joint internships for universities and companies, clarify intellectual property rules in advance, and allow researchers to experience industrial work without immediately severing ties with academia.
The connection needs to be two-way: not only transferring researchers to companies, but also introducing real problems, data, and engineering experience into the research environment.
The Innovation Authority is already operating and supporting programs designed to address the shortage of artificial intelligence experts. And connect advanced training with industry needs.
To have an impact on a national scale, a broader continuum of tracks is required—from a doctorate, through applied training, to integration into research and development teams.
Scientific depth as a competitive advantage
In the world of artificial intelligence and deep technology, advantage is not measured only by how fast you write code. It depends on the ability to identify an important problem, understand existing knowledge, propose a new way, and test whether it actually works.
These are precisely the skills that researchers acquire during their years of work in academia.
Israel has already invested a lot of resources in training these researchers. If it also provides them with applied tools, industrial experience, and a network of contacts, it can get more out of the investment — and strengthen both research and industry.
We don't have to choose between academia and high-tech. We need to build a better transition path between them.
Haim Geron is the founder and co-CEO of Infinity Labs R&D, which operates theAnd a foundation to train researchers for artificial intelligence positions in collaboration with the Innovation AuthorityThe article expresses his position.
More on the subject on the science website
- The Innovation Authority selects three academic knowledge commercialization accelerators for Israeli high-tech
- Yishai Frankel: Industry must adopt a model of cooperation with academia
- The national program for artificial intelligence was revealed
- Research: Almost all high-tech workers in Israel already use creative artificial intelligence tools
- More articles, fewer ideas? The danger of using artificial intelligence in science
