The beginnings of AI can be traced back to the 1950s and while AI has been improving for several decades, compared to other technologies, such as telephony, home computing and the internet, progress on AI has been relatively slow.
By Hans Roth, senior vice-president and GM: EMEA at Red Hat
Now it feels like the time when AI is available to almost everyone, and the technology industry is having a new moment. A moment in which we all have to ask ourselves how AI can help us do more with less. The application of AI to everyday tasks, such as writing, image generation, and music production, represents a watershed moment in terms of public consciousness.
In my conversations with businesses, I encourage them to look beyond the hype of AI and treat it as just another technology. In other words, as a tool to fix problems and seize opportunities. Just as cloud computing has answered the challenge of scalability, blockchains have addressed the issues of centralisation, and digital advertising software has ensured marketing budgets are spent more efficiently, AI should be judged by the outcomes it can generate.
So, what are these challenges and opportunities? At Red Hat, we have started to frame the AI conversation around seven key business realities.
* Efficiency versus innovation: In the pursuit of operational efficiency, businesses often find themselves compelled to accomplish more with less. Maximising output with limited headcount means harnessing existing skills better, by bridging knowledge gaps, developing new skills, and creating the conditions for frugal innovation.
* Managing complexity: The relentless march of software innovation promises boundless potential, yet it can also spawn intricate complexities. Each new system and integration comes with risks, such as security threats, service disruption or sudden surges in demand. The popularity of hybrid cloud computing can add to this burden. While event monitoring systems offer a level of control, IT teams can quickly be overrun by the sheer scale of their growing ecosystem and volumes of data.
* Enabling automation: Against the context of these first two realities, automation has become a key priority. Enabling automation affords staff to be freed from mundane tasks and redeployed to higher-value work. But automation raises questions of what to automate, with what tools, and how to trust that it’s working?
* Scaling to demand: Operating with limited resources is only one facet of the challenge facing IT teams. They are simultaneously tasked with scaling their operations to meet the soaring demand for applications and services. Keeping up with the demand for both DevOps and full-fledged production environments is not simply about enablement, but also how you then manage what you enable
* Connecting the edge: If the above points weren’t already enough of a challenge for IT, then enter edge computing to make life even harder. Data centres are no longer the sole hub for processing data. The edge is not just a different ‘place’ to do computing, but an entirely different approach. At the heart of the challenge is the conundrum of how to apply standards in data processing, accessibility and security to edge infrastructures and machines that are designed to be diverse.
* Balancing innovation with security: Unfettered innovation risks security; yet over-zealous security will stifle the will and means of creativity. Businesses must decide their own sweet spot on this spectrum, and continually adjust their operations and culture to fit with it. Embedding security capabilities and protocols within the software supply overcomes the view of security and innovation as a trade-off. Instead, this approach positions them as complementary functions, and instils developers with assurances and confidence in the safety of their work.
* Planning for sustainability: Governments, shareholders, customers, and employees are demanding organisations face up to their sustainability responsibilities like never before. It can create a mixed message for IT teams; on the one hand, do more, while on the other hand, preserve energy. Key is the ability to track and report sustainability insights and adapt working patterns to foster more sustainable practices.
AI serves as a versatile tool that can assist organisations in addressing these challenges. But what really links these seven realities is not just that AI can be applied to them all; but more importantly that AI alone is not enough. With each point, humans are the real secret weapon. Without people to identify, prioritise, engineer and evaluate the issues and the fixes, AI will at best have no impact; and at worse, result in profoundly negative consequences.
This is a key point I often urge executives to consider; that an AI application is only as good as the data it is trained on. Volume of data should not be a criteria. What really matters is focus; how relevant is the training data to the context of your organisation?
We call this ‘domain-specific AI’, and it presents a watershed moment in the evolution of AI. When an AI application is trained on private, targeted data and customised to the standards and practices of a specific business or industry, it has more ability to deliver truly unique and differentiated services.
Open source is by far the best option for building domain-specific AI solutions. Any open software benefits from a broader exchange of ideas and the collaboration of more talent. Indeed, pretty much every business AI tool that I can think of is an example of open source technology (yes, ChatGPT included). What I think confuses and concerns business leaders is a misunderstanding of ‘open source’. It is the codebase of the software (in this case the AI application) that is open and available for anyone to see and share. The data that it is trained on and generates, that’s as private as you want it to be.
Ultimately, AI’s true power lies not in its algorithms alone but in the synergies of human insight, human collaboration, data relevancy and computer processing. Executives who grasp this fundamental truth will very soon be able to claim they are at the forefront of something new.