‘India Must Capitalise On Its Own Strengths For AI Policy’

In light of AI supremacy discussions, comparisons between Israel and other nations arise, including India. Despite US and China's lead, there's potential for innovation. Israel's focus on AI and cybersecurity integration stands out. Could you discuss Israel's national AI policy and its role in broader strategies, especially in cybersecurity?

I think that every nation has to compete with its own tools. Israel will never have as much money as the US or as many people as China. But what we do really well is to do interdisciplinary work, connecting people together. What we also do very well is to take academic knowledge and transform that into startup companies. And so, the ecosystem here is very tightly intertwined into itself - with funding organisations, researchers, entrepreneurs, talent and government. This the security ingredient in our soup.

Regarding AI, we view it as a fundamental enabling technology, akin to the internet, mobile apps, and cloud computing. It spans various sectors such as health tech, transportation and food tech. Our strategy focuses on facilitating building blocks: talent, data, compute, testing environment and regulatory framework. By providing these preconditions, we enable spontaneous innovation across verticals, creating infrastructure for innovation to flourish.

According to numbers, there are about 2200 AI startups in Israel. What has helped this ecosystem grow and thrive fast?

While generative AI holds immense potential, numerous applications exist beyond it. Image processing, for instance, has achieved remarkable heights with AI, from medical imagery to infrastructure inspection and agricultural monitoring. Text and audio processing also show diverse applications. When discussing 2200 companies, it refers to those conducting R&D and us AI and machine learning across disciplines. For example, digital agronomy requires collaboration among agronomists, engineers, AI experts, operational leaders and CEOs. This brings it back to the secret sauce: bringing different people to the table together. That is what we do best in Israel.

Where does the government come into this picture? 

The key lies in multi-stakeholder consultation and collaboration to address market failures or barriers hindering successful AI ventures. Talent, compute and data are foundational requirements for AI, along with regulatory certainty. Governments may intervene to address shortages, such as in localised NLP, where commercial viability is challenged by limited language speakers. By providing critical infrastructure like NLP capabilities in local languages, such as Arabic and Hebrew, Israel aims to unleash innovation across various sectors. This infrastructure includes databases, pre-trained models and tools, empowering startups to focus on vertical-specific adaptations.

How are you approaching AI training? Are you looking at elemental training or is it specialised?

Israel is globally renowned for its abundant talent in AI. However, there remains a significant shortage, both in academia and the private sector. Each AI researcher in the private sector can catalyse the formation of an entire startup team, illustrating the critical role of expertise in fueling innovation. The conventional path to becoming an expert involves advanced academic training, including a master's or PhD, along with hands-on experience, spanning roughly 8 to 10 years. Yet, the demand for experts is immediate, prompting innovative solutions.

Israel's approach addresses this challenge through two main avenues. Firstly, leveraging the existing pool of PhDs in life sciences, who possess research skills and statistical acumen, to transition into data science roles. Upskilling or reskilling these individuals bridges the gap to becoming data scientists, creating a pipeline of highly trained experts. Secondly, Israel aims to attract foreign experts through streamlined visa processes and augmenting the local talent pool.

So, it is focus on post graduates and PhD-holders and inviting foreign experts to innovate and augment local talent…

While initiatives for basic AI training, such as bachelor's degree programs, are flourishing organically across cities like Tel Aviv, Jerusalem and Haifa, our targeted interventions focus on nurturing advanced expertise critical for driving startup innovation. By addressing the shortage of AI experts through strategic upskilling, reskilling and attracting foreign talent, Israel aims to bolster its position as a global AI powerhouse.

What must countries like India do with its AI policy?

For sure, the scale of problems in India are very much different from Israel. Each country must tackle technology and AI based on its unique strengths. India boasts a vast pool of talent and the market offers ample opportunities for experimentation. The country’s robust local market provides a solid foundation for startups to initially focus on domestic sales by leveraging cultural and linguistic alignment. Conversely, Israel, with its smaller market, must prioritise exporting from the outset.

My advice is simple: capitalise on your strengths. Instead of emulating others' strategies, tailor your approach to your country's unique advantages. A one-size-fits-all policy won't suffice; strategies must be customised to maximise each nation's strengths and opportunities.

How much has Israel invested on its AI infrastructure?

When discussing AI infrastructure, we take a holistic approach encompassing talent, compute, databases and regulatory environment. This infrastructure investment is ongoing, spanning the past two years and projected to continue for the next two to three years at least. We are committed to enhancing the talent pool from academia and industry, establishing accessible compute infrastructure for startups and researchers and ensuring the availability of essential data.

The entire program is budgeted at one billion shekels, approximately USD 250 million, allocated in two batches of half a billion shekels each. We anticipate a third batch within the next year or year and a half. This batch-oriented approach allows us to adapt to the dynamic nature of the field, focusing our efforts efficiently. The program commenced in early 2022 and is expected to run until 2026 or 2027.