AI isn’t a magic fix, but the right tools can bring us close
15 Jul 2024
6 mins
Artificial Intelligence (AI) is completely reshaping society; from bridging communication gaps across cultures and languages with AI-powered translation to revolutionizing healthcare by improving diagnostics through image recognition. According to a survey by McKinsey & Company, around 80% of executives believe that their companies could achieve a competitive advantage through the implementation of AI technologies, so it is not surprising that the adoption of AI has tripled in the last year within organizations across all industries.
However, AI cannot automatically add value and make a meaningful difference just because it has been implemented within an organization or team. One must consider how much AI-driven platforms are reliant on human intervention at specific contact points to ascertain how effective AI truly is when implemented into an organization. It is therefore important to explore how the combination of human intelligence and artificial intelligence at the right times can unlock AI’s true potential and implementation value.
When AI goes wrong
AI can undoubtedly benefit businesses in many ways through, for instance, the automation of administrative tasks. However, in order to reap the full benefits of AI, organisations must implement it efficiently and effectively. If not, AI may not meet the high expectations that businesses may have.
When AI is implemented inefficiently into business operations, it prevents them from gaining real value from AI. For instance, according to a survey conducted by IDC, 70% of AI projects fail to deliver, often due to an untrained workforce and insufficient comprehension of AI technologies. Additionally, according to a report by Gartner in 2020, around 85% of AI projects were expected to fail due to a lack of clear objectives, poor data quality and insufficient stakeholder alignment. This high failure rate highlights the pitfalls associated with implementing AI into business, and how many organisations struggle to effectively integrate AI solutions into their operations, often resulting in wasted resources and missed opportunities for efficiency gains.
The workforce needs to be properly trained and on-boarded onto AI in order to gain the most value from it. AI also must be implemented at the correct place, at the correct time, so that AI and humans can work together efficiently.
If we focus on the example of AI-powered translation, many organizations - particularly within complex industries such as pharmaceutical, financial, and legal - would greatly benefit from AI to translate large volumes of content and quickly turn around advisory literature on patient guidance. However, since AI-translation is not 100% accurate, this output also requires human review from qualified language specialists with subject-matter expertise.
With the above in mind, it is clear that this post-editing stage remains the critical gap between AI-translation output and translation that is trustworthy and fit for purpose, specifically for complex industries such as pharmaceutical, legal, or financial services. It is therefore important for humans to teach and train AI during the process, rather than at the review output stage.
If you would like to read more, visit AI Journal for the original article.
Denis Davies
All from Denis DaviesRelated Articles
