AI and the power of open-source software

 

AI clearly has many benefits for business, in cost, speed and quality. But it needs to be implemented in line with shifting enterprise priorities. These include: managing the move to the cloud; the need for better security against emerging technologies (including AI); the urgent requirement in some industries to address legacy IT and modernise apps; and the digitisation of workflows.

 

These shifting priorities are resulting in an increasingly complex technology environment. And complexity is a problem for business. It delays decision making and increases risk. Handled poorly, it harms customer loyalty and employee engagement. Ultimately, it damages bottom-line profits.

 

 

There is a pressing need for organisations to navigate complexity, and to do so with confidence, so that they can become agile, innovative and open to new possibilities.

 

Central to this is a collaborative approach to problem-solving. Rather than pretending we can solve every problem on our own, we should realise the benefits of working with partners, collaborating within and outside our organisations. This is especially true now that enterprise technology has become too complicated for anyone to navigate on their own. Success and progress in today’s rapidly changing technology environment require a community mindset.

 

One significant example of a community mindset is the use of open-source software. Millions of developers use open-source techniques: the well-known GitHub open-source software resource on its own has 56 million contributors.

 

The reason for this is quality. An open-source approach enables organisations to benefit from best-of-breed thinking (much wider than that available in any one organisation). Security, based on the far-reaching experience of the community, is stronger. And system integration is easier because you don’t have to develop code in isolation: there is always someone with a view across systems.

 

Use open-source techniques and you’ll never have to risk rapid innovation alone, no matter where you’re headed.

 

 

Open source is one of the most important drivers for AI and it is where most innovation is happening. Combine the muscle of the statistical algorithms used in AI with the insight, experience and energy of community wisdom, and you have something very powerful.

 

But despite its power, AI needs leaders with vision and robust management. It will never be sufficient to view the use of AI as an end in itself. AI must address real enterprise problems defined by tangible and realistic objectives.

It should also be technologically agnostic, meaning that it can run in any environment – on premises, or on public, private or hybrid cloud – and on any hardware platform. This is one area where the value of the open-source community really comes into its own: there will always be someone with insight into the circumstances you are operating under. In addition, open source facilitates the interoperability of standards.

 

Accessibility is also important. As well as being easy for end users to operate, AI should allow any organisation using it to have direct access to the model. This provides transparency so that people can see t heir data sources and better assess accuracy; it also means they can make adjustments to get better results for their business. One example of accessibility is seen with InstructLab, a pioneering platform from Red Hat and IBM that enables people to train AI models easily for their own use, while benefiting from the quality delivered by an architecture created by IBM and supported by the whole community.

 

The focus of any AI project should always be on the requirements of the end user organisation, not the options offered by their vendors. Open source helps democratise AI, by ensuring it is developed by diverse groups of people around the world, with specific input to address real world needs, often with smaller, domain-specific models rather than a one-size-fits-all large language model (LLM).

 

 

 

Collaboration across industry underpins the development of high-quality AI services. By using open-source methods, any business can start their journey towards effective AI in the knowledge that they will be harvesting the expertise and experience of thousands of individual developers and organisations.

 

You too can benefit from this community wisdom at Red Hat Summit: Connect events where you can network with your peers and learn about cutting-edge and enterprise-ready technologies based on open-source software. These are the go-to places to learn more about AI and Red Hat’s AI products.