Balancing the burden of AI democratisation – Silicon Republic

This post was originally published on this site.

Prashanth Ram discusses how modern organisations need to look inwards to address the challenges of large-scale AI implementation.

Implementing AI and other automation tools into workplace policy and practice is no mean feat. The resources required, for example time, money, energy and equipment, alongside the near constant training requirements can make it seem as though there is no clear end in sight. 

For Prashanth Ram, the CTO and co-founder of tech talent platform Smoothstack, the key to successful implementation is shared responsibility. This starts with democratising access to AI training, to prevent knowledge from becoming concentrated to technical teams and wealthy organisations. 

“The recent surge in generative AI exploration by companies represents a significant democratisation of AI capabilities,” he told SiliconRepublic.com. While traditional AI applications such as predictive analytics required specialised data scientists and engineers, generative AI has made sophisticated AI tools accessible to broader business users with varying technical expertise.

“The shift from complex model development to user-friendly interfaces has expanded AI’s practical business applications beyond traditional data science teams, enabling wider organisational adoption and new use cases that weren’t previously feasible for non-technical teams.”

But while there is evidence of a change in how AI access is shared, Ram said that challenges persist. For example many organisations struggle with ensuring data security and privacy, as it can be unclear how information can be safely processed by AI systems. Additionally, when it comes to governance and control, companies may lack the ability to establish effective oversight mechanisms to best monitor and audit AI interactions.

But primarily it is the growing skills gap when it comes to AI technologies that has the potential to delay, derail and even destroy AI democratisation at an early stage of organisational AI implementation. 

“Many organisations lack the foundational elements needed for AI implementation, such as data literacy across the workforce, modern data infrastructure and governance, understanding of AI capabilities and limitations among decision-makers, and change management expertise to handle AI-driven transformations.”

A problem shared, a problem halved

For Ram, the burden lies not with one person in a position of power, or one group, but rather, it is the responsibility of organisations as a whole, from the bottom all the way to the top, to ensure that efforts are taken to close the growing AI skills gap. A problem that is increasing in complexity due to the skills required becoming ever-more technical.

“Effective AI implementation requires a unique blend of skills, deep technical knowledge (machine learning, data science, programming), domain expertise (understanding specific business contexts) and infrastructure knowledge (cloud computing, security protocols). Finding professionals who possess this combination is challenging.”

Employers should create structured AI training programmes aligned with the companies’ specific needs, while also providing easy access to learning platforms, certification opportunities and pathways to further career development, Ram says. It is then the responsibility of the employee to take advantage of the chance to upskill in AI and learn more than just the basics. 

This can be through participating in new training opportunities, engaging with online learning, identifying how AI might affect or enhance individual roles, applying what has been learned to the job, and staying informed on AI developments in the industry. By knowledge sharing with co-workers you can also work on your soft skills whilst contributing to the wider workplace conversation on AI advancements. 

“Most successful organisations are finding that simply expecting employees to upskill independently isn’t effective, it requires a structured, supported approach with clear investment from leadership,” Ram said. “This ensures consistent skill development aligned with organisational needs rather than scattered individual efforts.”

He advises organisations to pay attention to skills beyond what is considered standard AI knowledge, in areas such as data literacy, problem decomposition, structured thinking, business process analysis, written communication and project management. Domain specific knowledge, for example industry regulations and compliance requirements, business workflows and operational constraints, are also important. 

For Ram, democratising AI training access is crucial for reducing inequality and workplace silos, empowering workplace transformation, limiting the potential risks of advanced technologies and enabling organisations to have a much wider economic impact. It also creates resilience, so organisations skilled in AI on a broad and individual level, are better positioned to adapt to technological changes.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.