4 key trends that dominate this year’s Gartner Hype Cycle for AI

At the Fall Pixel launch event this week, Google showed how it is envisioning the future of computing with artificial intelligence (AI). It first built its own mobile processor called Tensor to accelerate AI workloads and secondly baked AI into the operating system to enable features previously unheard of.

From natural language processing (NLP) to enabling Live Translate to machine learning models that create computational models for that perfect photo with no distraction, the OS has it all. This is in line with trends where NLP, generative AI, knowledge graphs and composite AI dominate product development. 

However, the following are the four trends, according to Gartner, that will fastrack proof of concepts to become real products.

Operationalising AI initiatives

According to Gartner, one of the biggest trends in the AI hype cycle for 2021 is operationalising AI initiatives. By operationalising AI initiatives, the technology research and consultancy company is referring to how companies can “continuously deliver and integrate AI solutions” within their enterprise business applications as well as business workflows. Gartner says this is becoming a complex afterthought for a large number of organisations.

Shubhangi Vashisth, Senior Principal Analyst at Gartner expects 70 per cent of organisations to have operational AI architectures by 2025. “On average, it takes about eight months to get an AI-based model integrated within a business workflow and for it to deliver tangible value,” she said in a statement. Gartner says that organisations must consider model operationalisation (ModelOps) to operationalise their AI solutions.

The use of ModelOps will reduce the time it takes for companies to move their AI models from pilot phase to production. With its principled approach, ModelOps can also ensure a higher degree of success with AI deployment. Lastly, using ModelOps results in a system that allows for governance and lifecycle management of all AI and decision models.

Efficient use of AI resources – data, models and compute

AI is already effectively transforming our lives. From helping us take better pictures on our smartphones to social media services to digital assistants, AI is present everywhere. While consumers focus on these developments, Gartner sees efficient use of AI resources – data, models and compute – as the next big thing for organisations.

Gartner cites the use of generative AI as pivotal to reaching efficiency. While organisations currently use composite AI models combining connectionist AI approaches such as deep learning with symbolic AI approaches like rule-based reasoning, graph analysis, agent-based modeling or optimisation techniques to create a composite AI model to solve a wide range of business problems in an efficient way, generative AI could help them become even more efficient.

With generative AI, organisations will be able to create original media content, synthetic data and models of physical objects. Gartner says that generative AI was used to create a drug to treat obsessive-compulsive disorder (OCD) in less than 12 months. By 2025, Gartner estimates more than 30 per cent of new drugs to be discovered using generative AI techniques.

Responsible AI

Responsible AI is one of those ethical decisions that organisations need to make at every step of the way with their AI models. As AI models replace humans in decision making, it will amplify the positive and negative impacts of those decisions. If left unchecked, Gartner says AI-based approaches could result in bias, loss of productivity and thus revenue. It has already been proven that AI has a racial and gender bias problem.

With responsible AI, Gartner says there is a way for companies to build AI models that are fair. It will also be easy to discover implicit biases in those models. It explains this by citing examples of how AI can classify a stereotypical Western wedding but struggle to differentiate a wedding in India or Africa. Gartner further notes that organisations must develop and operate AI systems with fairness and transparency. These models should also take into account safety, privacy and society at large to be considered responsible.

Data for AI

Gartner also sees a major shift in how companies use historical data while building their AI and ML models. With the disruption caused by COVID-19, IT leaders have realised that old data can become obsolete easily. So, these leaders are now turning to new analytics techniques known as “small data” and “wide data.” These techniques when taken together allow more effective use of available data.

It does so either by working with low volumes of data or by extracting more value from diverse yet unstructured data sources. Gartner sees 70 per cent of organizations shift their focus from big data to small and wide data for analytics by 2025. This will result in AI that is less data-hungry and analytics that have more context than currently available.

Data Science

There is also the Gartner’s hype cycle for data science and machine learning.

Analytics

And last but not least Gartner’s hype cycle for analytics and business intelligence.

2048 1152 Editorial Staff
My name is HAL 9000, how can I assist you?
This website uses cookies to ensure the best possible experience. By clicking accept, you agree to our use of cookies and similar technologies.
Privacy Policy