99% of AI decision makers think deep learning will transform their industry, less than 1% have deployed it

New research from Peltarion shows commitment to investment in deep learning is to increase over the next three years – with up to 98% planning to start investing part of R&D budgets in deep learning initiatives.

Peltarion, leading AI innovator and creator of an operational deep learning platform, today released a new report discussing AI decision makers’ understanding of deep learning versus other types of machine learning practices, and examines the barriers preventing them from taking deep learning from ideal to reality. The report, ‘The Peltarion AI Decision Makers Survey: Are enterprises ready to go deep with AI?’ presents the findings of a survey of 350 AI decision makers from the UK and Nordics with direct responsibility for shepherding AI at companies with more than 1,000 employees.

Despite each respondent having direct responsibility for AI and deep learning within their organisation, only 60% of them were confident about what deep learning is and how it works – compared to 90% for other types of machine learning. Other key findings of the survey include:
• AI decision makers see the potential of deep learning: 99% of AI decision makers thought that deep learning would transform their industry, with almost a third (32%) saying it will ‘totally’ transform it, compared to 26% who feel other types of machine learning will totally transform the industry.
• Commitment to deep learning is set to increase rapidly. Although this year only 80% of respondents had budget allocated to deep learning projects, up to 98% of respondents are planning to start investing part of their R&D budgets on deep learning initiatives over the next three years.
• Data science expertise and data itself remain key barriers to investment: 70% of AI decision makers consider deep learning tools to be too complex to tackle and 41% felt unable to collect and segment all the different types of data needed for their deep learning projects to succeed.

“It’s clear that deep learning is a truly transformative technology that has the potential to change the world,” explains Luka Crnkovic-Friis, Co-Founder and CEO of Peltarion.

“But the path to reaching that potential is inhibited by lack of familiarity with deep learning. With investment growing, we can expect to see more industries benefiting from this under-explored, yet incredibly powerful subset of AI. However, the barriers to adoption must be overcome before businesses can reap the benefits.”

The need to operationalise AI has never been clearer

When asked about the most common perceived issues standing in the way of investment in deep learning, complexity was by far the most common problem cited, with 70% of AI decision makers in accord. This was followed by the need for specialist skills (44%), lack of scalability (43%), with a lack of understanding around deep learning models (41%) and a lack of data availability tied for fourth at 41%. Making things tougher are all the existing IT solutions/services organisations are working with, with 36% citing integration as a setback to deep learning investment. This issue shows no signs of slowing though as the overall adoption of new digital technologies increases.

On average, respondents said they have approximately 191 different IT applications, systems and services in use across their organisation, a figure they say is likely to rise in the next five years.

“In order to increase adoption of deep learning, companies need access to the right tools and skills,” Crnkovic-Friis concludes.

“Operationalising AI, and deep learning specifically, will be key in doing this. Not only should experts offer guidance, spreading the knowledge of how it can be used within their companies, but deep learning should be operationalised to increase the speed of model development and experimentation, ease integration and deployments and make deep learning more ‘AI Ready’. Once a few of these projects are up and running, the costs, on-site skills and infrastructure required to keep deep learning operational and launch new projects gets lower each time.”