We have officially entered a new stage of artificial intelligence (AI). This technology has sparked people’s curiosity about what’s possible since the 1950s with the creation of the initial rule-based systems. Fast forward 70 years, generative AI (GenAI) is the accelerator of the decade, representing a significant and immediate opportunity to fundamentally change how companies operate. Its impact on the world’s economy could amount to between $2.6 trillion and $4.4 trillion annually.
The noise around GenAI has been so loud, and the technology has as many advocates as sceptics. So, I want to go back to the basics of this emerging technology: where GenAI currently stands, its benefits, and the challenges we must overcome to embrace it.
Simply GenAI
GenAI shares similar characteristics with traditional AI, most notably turning data into algorithms. But the two are quite distinct in terms of results. While traditional AI aims to generate predictions based on input information, GenAI produces various forms of content, such as text, images, audio, and video, creating fresh insights based on vast data sets.
The idea of leveraging data to create insights has been around for decades in the IT sector. Still, the latest powerful capabilities and people’s continuous passion for content creation have helped GenAI gain significant momentum in recent times. This phenomenon may explain why we all have been fascinated by ChatGPT, whose launch made the GenAI phenomenon simply inescapable.
The Benefits
Despite many of its transformative capabilities remaining untapped, organisations across all sectors will come to rely on AI to bring better business outcomes. For example, it will increase employee productivity by picking up a significant proportion of day-to-day activities and automating mundane tasks, allowing employees to focus on more creative projects and higher-value activities. This will prove transformational in verticals like healthcare, for example, where GenAI can quickly summarise patient information, allowing nurses and doctors to spend less time reading notes and more time providing frontline care.
In addition, with so much data input, the emerging technology will produce better insights to improve business decision-making. Many are already using the idea generation process to obtain practical and/or innovative solutions to existing problems, opening new revenue streams in markets business leaders would not have considered entering.
65% of organisations that have moved beyond pilot stages expect to see meaningful results from their GenAI initiatives in the next 12 months
You cannot talk about GenAI without someone mentioning AI-augmented chatbots; conversational interfaces offering personalised and timely replies designed to boost customer satisfaction. The jury may be out on how successful Chatbots are at truly satisfying customers. But with the power to draw on immense data sets comes the vast potential to serve customers better – for example, GenAI’s multilingual ability means a business can reach and serve a diverse group of customers without relying on teams of polyglots.
The Challenges
It can be challenging for businesses to determine how best to generate value from GenAI while simultaneously navigating anxieties from the board, the C-suite, or employees on the challenges of this transformational technology. Our recent GenAI Pulse Survey revealed that 75% of UK IT decision makers (ITDMs) think the impact of GenAI will be significant for their organisations, but 49% are somewhat or very hesitant to adopt GenAI. And it’s understandable. GenAI doesn’t have to be complex, but there is much to consider to get it right.
For example, AI-generated content that is practically indistinguishable from human effort sounds exciting, but it also leaves us with a question about the owner of the content. What data to keep, how to use it, which use cases to apply it to, and how ubiquitous that access is across an organisation determines the extent of value it delivers. Data’s value is based on many factors, including volume, uniqueness, sensitivity, veracity, velocity, and usefulness.
It can be challenging for businesses to determine how best to generate value from GenAI while simultaneously navigating anxieties from the board
Skills may be a challenge for many, as the perfect mix of scientific and creative talent is necessary to understand the complexities and data requirements of GenAI models.
GenAI is only as good as the data it draws from, so businesses should be aware that any public models they use could be trained using incomplete, outdated, or biased data and algorithms, making any output less reliable. Additionally, since GenAI training models require a large amount of data, businesses must guarantee that the information is clean, well-labelled and illustrative of the problem to be resolved. GenAI outputs that impact human life should be meticulously examined to ensure they are safe.
GenAI is not independent of other architectures. Therefore, enterprises must ensure solid foundations before implementing generative AI safely and successfully. That means establishing a robust data strategy, investing in AI-ready infrastructure, and prioritising data privacy and security. In our recent Global Data Protection Index, 78% of UK respondents agreed that GenAI will amplify the specific value of data types, which will require additional and higher security measures. This explains why UK respondents highlighted the ability to protect multi-workload environments (61%) and ensuring cybersecurity across all multicloud environments (63%) as the capabilities most important to their organisation when enabling hybrid, multicloud operations. Whether in on-premises data centres, in the cloud or at the edge, GenAI models need significant computational capability to run, and substantial data protection. Organisations managing large data sets without the foundations in place may find GenAI adoption quite onerous.
When it comes to the data that feeds your GenAI models, the best approach for most organisations will be to bring AI to their data vs. the alternative of moving it to a cloud-based solution. Considerations such as data gravity, availability of models, data privacy, security of enterprise IP, lifecycle management and cost of cloud storage and computing will all impact the performance and time-to-results of GenAI projects.
Lastly, businesses must grasp the GenAI implications around regulatory compliance, content ownership, and ethics.
What the future holds
Until now, we’ve seen much experimentation around GenAI but little in real-world, scaled GenAI activities. That will change in 2024. GenAI projects will shift from proof of concept to proof of productivity gains with greater adoption and scale across organisations and industries.
According to our GenAI Pulse survey, 65% of organisations that have moved beyond pilot stages expect to see meaningful results from their GenAI initiatives in the next 12 months. Those innovators will reap the first mover’s advantage, with GenAI becoming mainstream in most organisations in the coming years.
With that in mind, businesses must focus their GenAI efforts to achieve the most significant ROI. We have witnessed the positives, and we know where challenges lie. Now is the time that we’ll see enterprises shift from broad experimentation to a top-down strategic focus, picking the select GenAI projects that will be truly transformational.
