AI can amplify Kazakhstan’s global voice, says strategist Olaf Groth

Global investment in AI is at an all-time high, yet many markets are struggling to see a clear return on investment. In your view, what usually prevents these massive capital flows from turning into practical, everyday benefits for businesses and citizens?

The biggest obstacle to achieving a return on investment is the human and organizational factor. We are failing to transform organizations as quickly as technology develops. You could also say that technology is developing too quickly for people and organizations to adapt. Institutions, including rules, customs, legal frameworks, processes, procedures and structures, change very slowly.

We are having trouble keeping pace. We need to rethink how we communicate, make decisions and build things, as well as the processes and structures that support those activities. That takes time.

We need more leaders who focus on managing change and organizational transformation. We also need to become better at experimenting thoughtfully. That means understanding why we are experimenting, how we conduct experiments and how we measure the results. We need to combine open-minded discovery and blue-sky thinking with ways to capture and assess what we find.

When experiments produce positive results, we need to put them on a pathway toward a return on investment. People are sometimes reluctant to measure what they are experimenting with, but you can be both free-spirited and disciplined about measuring your discoveries.

These transformations take time, and we are now in a race against time because we are on a pathway toward spending $10 trillion over the next couple of years. That investment needs to generate returns. Otherwise, those providing the capital may decide to slow down or stop funding, which would be the worst outcome.

You’ve been tracking Central Asia’s momentum in the digital space. Looking at similar emerging markets around the world, what unique strengths or advantages can new hubs like Kazakhstan leverage to hold their own against established global AI players?

Kazakhstan has already done some remarkable things. I helped introduce Minister Zhaslan Madiyev to OpenAI, and they were very interested in what Kazakhstan had already built. I believe you had developed a model with 180 billion parameters in Kazakh, Russian and English. OpenAI was interested because that could allow a partner like OpenAI to train its own models in your languages.

That is a good start, but Kazakhstan has much more to offer. It has tremendous natural resource wealth. It is an agricultural nation that can feed up to 400 million people. For much of its recent history, it has also been a logistics hub and a transit route for trade.

Think about the space economy and Baikonur as well. Kazakhstan needs to bring those assets to its partners because the future is physical AI. AI will extend beyond digital systems and text-based understanding to interacting with the real world, including the natural and built environments.

Kazakhstan can bring its natural, industrial and economic assets to AI innovators and become a valuable partner.

It sits at a geographic and geoeconomic intersection that is tremendously valuable. Think of it as a switchboard. Historically, that switchboard has been largely analog. Now it can become digital and, ultimately, a cognitive, AI-driven switchboard.

Globally, there’s a debate between strict regulation and aggressive tech expansion. How can growing digital economies build data rules that protect their national sovereignty and citizens without slowing down innovation?

The best way to make data more productive is to create a regulated, safeguarded and accessible market that connects different groups of data providers and users.

Most of our data today is locked in silos behind corporate or government walls. There is little reason for corporations to retain all that data when much of it has nothing to do with the services they provide.

We should leave some room for unexpected discoveries. But, broadly speaking, we need to make data available through a market where certified traders and analysts can see what exists and assess its value. We must protect privacy and ensure that people retain control over how their data is used.

We also need to assign data a value, price it and match supply with demand. That would create an efficient market, similar to a capital or commodities market, where people can find what they need and pay a fair, market-based price. I see no reason why we should not do that with data, provided we ensure security, safety, privacy and individual control.

The next generation of solutions will involve people and institutions having their own AI agents. These agents would understand what data an individual creates and what information that person needs to become more productive, healthier or better informed.

In the future, this could involve one agent or a coordinated group of agents. Some would carry your personal or organizational data privacy preferences, telling the market what you are willing to disclose and what you want to keep private.

That is the emerging world of the modern global data economy. Countries are experimenting with these approaches, but we are not there yet. People are accustomed to data being either free or unavailable, which is one reason there is so much demand for synthetic data. Instead, we should make the data we already have available with the right safeguards.

Kazakhstan could lead in creating such a market, just as the UAE has taken a leading role in regulating cryptocurrency. People look to the UAE as a place to experiment with new rules and frameworks. That is a valuable role for a hub such as the UAE or Kazakhstan.

My hope is that Kazakhstan can create a safe, secure data market that protects people’s control over their data and becomes a global hub.

You mentioned the importance of safeguards. Should powerful AI systems have an “off switch” that allows them to be shut down when necessary, and should similar controls apply to data transfers?

Absolutely. There should be an off switch for both data transfers and AI systems. There also need to be clear, substantial penalties for violations.

We need better regulation. It is a fallacy to think that regulation is always bad or that it necessarily holds business back. Smart regulation can enable business. President Tokayev made that point in his speech, and I thought it was very enlightened.

We cannot rely solely on self-regulation. We can start with it, but we need to build on it with rules that are business-friendly while strictly protecting individuals’ dignity and well-being. When violations or wrongdoing occur, there must be substantial penalties.

Markets reward clear rules. Businesspeople value them. What they dislike is uncertainty that prevents them from assessing risks and opportunities.

Founders in fast-growing tech hubs often hit a wall when trying to expand internationally. Aside from the obvious need for funding, what are the biggest structural hurdles regional AI talent faces on the global stage today?

There are several bottlenecks. First, Kazakhstan needs to broaden and deepen its capital markets. They are currently too shallow. There is not enough capital available on appropriate terms across the different types of investment. You need more private equity and venture capital, so capital markets must be a priority alongside data markets.

The two need to develop together. That requires clearer rules that encourage private capital participation. The government should safeguard capital markets and can participate in areas such as the bond market, but it should not dominate them overall.

Second, entrepreneurs need mentors. Third, they need opportunities to scale internationally, across Central Asia and into Europe, the United States and potentially China.

That requires multilateral agreements, or agreements among smaller groups of countries, that create digital free trade zones. These would allow Kazakh entrepreneurs to take their capital, ideas and intellectual property into international markets.

But money, data, expertise and a venture are not enough without mentors who can advise entrepreneurs on what to do and what to avoid. Kazakhstan needs to attract international mentors and help its entrepreneurs enter foreign markets with the right coaching and support.

That is what a good venture capital investor does in the United States: provides funding, opens doors and coaches founders. Kazakhstan needs more of that.

Fourth, entrepreneurs need help navigating the geopolitics of AI. Kazakhstan is particularly interesting because it sits at the intersection of China, Russia, Europe, South Asia and the Middle East, with the United States somewhat farther away. It has a strong relationship with the United States and should develop it further, while also recognizing the realities of its geography.

The question is how to build ventures at these technological and geopolitical crossroads. That position provides access to many different assets, but it also requires geopolitical risk management and mitigation. The aim should be to turn those risks into opportunities.

How do you work with Chinese and American innovators in the same data center or economy? We can develop monitoring and communication protocols that help manage crises and prevent misunderstandings.

This combination of safeguards and opportunities is particularly relevant to Kazakhstan because of its position at these intersections.

Halford Mackinder’s Heartland theory emphasized the strategic importance of Eurasia’s interior. Do you see a modern parallel in Kazakhstan’s potential to become a hub for digital connectivity and AI, turning its geographic position into greater economic and geopolitical influence?

The British recognized the importance of Central Asia. European Commission President Ursula von der Leyen also recognized it recently when she met, I believe, with President Tokayev. She said she believed Kazakhstan would become a major hub in the global economy, and I share that view.

As I said, Kazakhstan could become a hub comparable to the UAE or Singapore. In that sense, it can become an intelligent hub that is neutral, works well with others and has its own voice.

That last point is very important. To have influence, you need your own voice. The idea of the Heartland determining its own path and influencing geoeconomic history depends on building a strong position that is both defensible and attractive to others.

That is why I say Kazakhstan should build on the strengths it already has: STEM talent, space economy assets, natural resources, uranium, logistics routes, and now data, AI experimentation and the education platform Minister Sayasat Nurbek is putting in place, which is very impressive.

Together, these can become sources of technological, scientific and economic influence. You are creating value by bringing all these assets together.

Although Kazakhstan has a relatively small population, it can be powerful in terms of the creativity generated here, particularly through research and development connecting AI with these other fields.

Today, major powers are increasingly protective of their tech, and global Big Tech firms dominate the market. How can mid-sized countries build strong partnerships with foreign tech leaders while keeping full control over their own digital future?

We used to say that the major technology companies, perhaps a dozen across the United States and China, were creating oligopolies or structures resembling them. It seemed very difficult for smaller innovators to compete, especially when large platforms acquired innovative companies and absorbed them into their own organizations.

But look at what has happened with OpenAI, Anthropic and Mistral. You could now arguably include Higgsfield. These are specialized companies with strong, defensible positions that are giving the larger hyperscalers serious competition.

In the same way, I do not think Kazakhstan’s smaller size or more limited financial resources compared with Silicon Valley or Shenzhen need to be decisive.

What matters is making use of your assets and maintaining global connections. Send your people into international markets, and then bring back human and financial capital, intellectual property and market knowledge.

I see tremendous potential for AI to amplify Kazakhstan’s voice and extend its reach. It is a two-way relationship: Kazakhstan will benefit, and so will the world.

Earlier, Qazinform News Agency reported that NVIDIA is set to participate in the next phase of Kazakhstan’s Data Center Valley project in Ekibastuz. During a meeting with company representatives, President Kassym-Jomart Tokayev emphasized the importance of expanding AI use across priority economic sectors. Two supercomputing clusters based on NVIDIA technologies have already been launched in Kazakhstan.