How to ensure that AI projects do not fail
If you have noticed an uptick in product recommendations based on your Amazon purchases, or GPS services that are increasingly accurate in displaying congested traffic areas, it is because artificial intelligence is everywhere. Adoption of artificial intelligence in organisations have tripled in the past year, and artificial intelligence is a top priority for CIOs.
Yet early artificial intelligence initiatives have a high probability of failure due to misalignment with business requirements and lack of agility.
Although the potential for success is enormous, delivering business impact from artificial intelligence initiatives takes much longer than anticipated. IT leaders should plan early and use agile techniques to increase relevance and success rates.
Gartner predicts five things that CIOs should consider in the rapid evolution of artificial intelligence tools and techniques, and how they will play out in their organisation.
Infrastructure decisions
The use of artificial intelligence across enterprises is ramping up quickly. In fact, through 2023, artificial intelligence will be one of the top workloads that drive infrastructure decisions. Accelerating artificial intelligence adoption requires specific infrastructure resources that can grow and evolve alongside technology. Artificial intelligence models will need to be periodically refined by the enterprise IT team to ensure high success rates.
Manage complexity
One of the top technology challenges in leveraging artificial intelligence techniques like machine learning or deep neural networks in edge and IoT environments is the complexity of data and analytics. Traditional artificial intelligence use cases that do not involve customer expectations are successful because of the tight collaboration between the business and IT functions, so securing the help of internal engineering teams is a must.
Simple techniques
Through 2022, over 75% of organisations will use DNNs for use cases that could be addressed using classical machine learning techniques. Classical machine learning techniques are extremely underrated. Once you sift through the artificial intelligence hype, you will realise that many organisations are pushing to apply deep learning techniques without even understanding how they apply to their current initiatives. As such, simplicity is key, and IT leaders should take the time to learn the spectrum of options to appropriately address their business problems.
Serverless computing
Containers and serverless computing will enable machine learning models to serve as independent functions and, in turn, run more cost-effectively with low overhead. A serverless programming model is particularly appealing in public cloud environments because of its quick scalability, but IT leaders should identify existing machine learning projects that can benefit from these new computing capabilities.
Adopt automation
As the amount of data that organisations have to manage increases, so too will the abundance of false alarms and ineffective problem prioritisation. With the shortage of digital dexterity talent in I&O to effectively adopt artificial intelligence, automation is a key solution. By 2023, 40% of I&O teams will use artificial intelligence-augmented automation in large enterprises, resulting in higher IT productivity with greater agility and scalability.
Key takeaways
- Early artificial intelligence initiatives have high probability of failure due to misalignment with business requirements.
- Delivering business impact from artificial intelligence initiatives takes much longer than anticipated.
- IT leaders should plan early and use agile techniques to increase relevance and success rates.
Chirag Dekate at Gartner explains that AI projects tend to fail and provides five basic pointers to help ensure their success in an organisation.
