A lot has been written about how Artificial Intelligence is coming to our workplace. How productivity boosts and customer engagement will deliver bottom line returns to justify the capital spend. It’s a seductive pitch for the employer. So where are organizations focused on adopting Artificial Intelligence into their organisation?
Creating and sustaining competitive advantage are primary drivers as are deriving returns on investment in cognitive technologies from better operations and customer service. Dresner Advisory’s recent study highlighted Finance, Marketing and Sales departments are the most proactive in exploring and piloting AI initiatives. Business intelligence functions and Recommendation engines being common projects. R&D departments are focusing on data integration and predictive analytics. It’s the shared interests between departments that encourage organisations to pursue cognitive technology initiatives.
Reilly’s AI adoption in the Enterprise report echoed these findings;
Regardless of use cases and technologies being adopted, the current adoption of AI seems to revolve around manipulation of four categories of data;
- Audio (including speech)
- Computer Vision (eg. images and video)
- Structured Data (eg. time series, logs and geo-spatial)
Healthcare organizations are driving computer vision applications such as radiology image analysis whilst other sectors focus their deep learning efforts on structured data.
Deloitte’s State of AI in Enterprise report demonstrates changing expectations on what AI will offer to their organisations. Deloitte’s respondents are realising the need to enact operational changes before being able to see the benefits AI can bring.
Despite the high level of intention to spend more on AI technologies by executives surveyed; this trend should be viewed with caution given the lack of expected rigour the spend is being tracked for return on investment. When more mature technologies are brought in to be trialed, established metrics to track the success of the project are employed. The lack of these being used suggests an uncertainty on expected returns and difficulty in justifying continued adoption to the Board.
Why isn’t it happening faster?
Many organizations just aren’t ready to replace jobs with AI without structural change to their business. Deloitte identified key operational challenges including;
- Data access, integration and privacy assuming it was available in the first place.
- Implementation and onboarding.
- Integrating AI capabilities into existing business roles and functions.
- Justifying cost of developing AI solutions.
- Lack of available human talent internally or through recruitment.
- Measuring and proving business value.
The early adopters’ challenges of embracing AI are also covered by Reilly’s 2019 ebook and summarised in the chart below. The leading bottlenecks are consistent with the need to drive internal operational changes in readiness for successful AI adoption.
Separate to operational challenges, many organisations highlighted that cybersecurity concerns and reliability of any AI system were primary reasons for balking. Despite the strong support from Finance and Marketing departments to rollout AI based projects; organizations flagged key risks such as predictability of outcomes, transparency and bias.
It appears Key Risk Indicators are dictating adoption rates more than Key Performance Indicators for now.
Deloitte suggests to early AI adopters;
- Apply the same operational discipline when executing AI strategies as when implementing more mature technologies.
- Appoint AI champions
- Process to drive prototypes to production
- Implementation roadmaps
- Identify relevant metrics, measure costs and returns on investment
- Consider Cloud based AI services as a way to achieve quick wins and momentum with lower upfront investment. Whether the organisation is prepared to expose sensitive data to a public cloud environment needs to be managed.
- Incorporate cybersecurity as a high priority from the beginning.
Deloitte concludes that a core decision around adopting AI is knowing where they want to “Automate to Replace” and where to build “Intelligence to Augment”.
At Verbz, we’re building to Augment the capabilities of our users and their teams. We believe there’s enormous untapped productivity and time saving in giving everyone the ability to action their ideas and intentions in the moment. It’s like having a superpower that gives you back time otherwise lost in catch up at the end of the day. We’re making it easy to act on thoughts and respond to tasks when it suits you so they’re not lost in the rush.