Open AI access will make a once-exclusive tool empowering for all

Prof. Syed Munir Khasru

South China Morning Post
March 19, 2025

Link: https://www.scmp.com/opinion/world-opinion/article/3302428/open-ai-access-will-make-once-exclusive-tool-empowering-all

Artificial intelligence (AI) has made quite the splash this year. Despite the many benefits its advancement can bring, a world where AI remains in the hands of a few elite corporations and nations will pose significant risks.

Without democratisation, only a limited number of players will be able to afford AI-driven solutions. This could lead to increased economic inequality, with wealthy organisations gaining disproportionate advantages in productivity and efficiency while smaller businesses and underprivileged communities fall further behind.

In such a world, AI’s potential to address global challenges, such as climate change, healthcare and education, would also be severely constrained, with solutions driven primarily by profit rather than societal good.

A centralised AI landscape could also exacerbate ethical concerns, with unchecked monopolies controlling vast amounts of data and decision-making power.

Ensuring AI’s accessibility is not just an economic necessity but a safeguard against deepening technological divides and global inequalities. Thus, that AI development is increasingly affordable and accessible represents a pivotal shift. With significantly reduced costs, AI is no longer an exclusive tool of major corporations but potentially a catalyst for global innovation, economic growth and problem-solving.

One of the most striking recent developments in AI democratisation is DeepSeek’s open-source R1 large-language model. While companies like OpenAI reportedly spent between US$80 million and US$100 million training their advanced GPT-4 model, DeepSeek managed to achieve impressive results with just US$6 million and 2,000 Nvidia graphics processing units (GPUs).

This dramatic reduction in cost signifies a transformative shift, proving powerful AI models are no longer confined to organisations with deep pockets.

The implications are profound: start-ups, researchers and educational institutions can now develop and train sophisticated AI models without requiring exorbitant resources. This shift not only diversifies the players in AI development but also fosters a more inclusive technological landscape where innovative solutions can emerge from different corners of the world, rather than being concentrated in Silicon Valley or Beijing.

A recent study by the Massachusetts Institute of Technology (MIT) observed the work of more than 1,000 materials scientists and found that those who used AI-powered tools saw a productivity increase. These researchers discovered 44 per cent more materials, which led to a 39 per cent rise in patent filings and a 17 per cent boost in downstream product innovation. The statistics illustrate AI’s ability to amplify creativity and efficiency, helping researchers and engineers solve complex problems at an unprecedented pace.

Industries ranging from healthcare and finance to agriculture and education stand to benefit from such AI-driven efficiencies. AI-assisted drug discovery, for instance, can accelerate the development of life-saving medicines, while AI-powered financial analysis tools can enhance investment strategies for individuals and small businesses.

Democratised AI is breaking down these barriers. One of the most promising examples is California’s partnership with Nvidia to train more than 100,000 students, faculty, developers and data scientists. By integrating AI education into community colleges and state universities, this initiative ensures AI literacy is not confined to elite institutions like MIT or Stanford but reaches students from diverse socioeconomic backgrounds.

As automation reshapes the job market, equipping people with AI skills ensures workers remain relevant, adaptable and capable of leveraging AI to enhance their careers rather than being displaced by it. Governments and educational institutions worldwide should take note and implement similar programmes to make AI education widely available.

Traditional AI development required vast capital investments, but with affordable open-source models, small businesses and start-ups in regions like South Asia, Africa and Latin America can now integrate AI into their operations.

For example, AI-driven agricultural analytics can help small farmers optimise crop yields with predictive weather modelling, while AI-powered translation services can enable businesses to expand into global markets. In the finance sector, AI-driven microfinance solutions can help underbanked populations gain access to credit, fostering economic growth from the ground up.

The key takeaway is that by providing open access to AI, we enable a broader spectrum of innovators to contribute solutions to pressing global challenges like climate change, poverty and healthcare disparities.

The advancements in AI affordability, accessibility and education point to a future where AI is not just a privilege but a tool that enhances global productivity and innovation. If nurtured properly, democratised AI can empower start-ups and small businesses with AI-driven insights and automation. It can accelerate scientific discovery by enabling researchers to solve complex problems faster. Additionally, AI can expand educational opportunities, ensuring AI skills are not limited to a select few.

Democratised AI can support economic development by making AI-driven solutions accessible to emerging markets and enhance global competitiveness, allowing diverse players to contribute to AI advancements. By recognising AI’s potential beyond its risks, policymakers, educators and business leaders can ensure AI remains an engine of inclusive progress rather than an instrument of control by a select few.

The democratisation of AI is not just a technological revolution – it is an economic and societal imperative. From increasing productivity in research to training the workforce of the future, AI’s impact will be felt across society.

To harness AI’s full potential, we must advocate for governance that supports open innovation rather than restricts it. Thoughtful policies, widespread AI education and continued investment in accessible AI technologies will ensure the benefits of AI reach everyone. If we succeed, AI will not be a tool of inequality but one of empowerment, driving a future where innovation knows no boundaries and progress is truly shared across the world.