How to identify tasks that will be replaced by generative ai

By
Curvestone
2
min read
Share this post

Determining which jobs and tasks will be automated by algorithms like GPT-4 can be a complex task. However, a simple rule of thumb can help you make an initial assessment. If you find that when you ask Chat GPT to perform a task, such as drafting, answering questions, or categorising, it completes about 80% of the task accurately, but the remaining 20% requires human intervention, it's likely that these jobs or tasks will be fully automated by language models (LLMs) in the near future. While it's challenging to provide precise predictions, a timeframe of 1-3 years seems reasonable.

At our company, we closely observe how engineers utilise the 80/20 use case, particularly those requested by our clients, predominantly Law Firms and Insurance Companies. We have witnessed significant optimisation of these 80/20 use cases, often within a matter of days to weeks. Therefore, it's only a matter of time before other companies recognise this potential and begin implementing LLM-powered platforms, workflows, bots, and apps to automate similar tasks and jobs.

Initially, automation was commonly associated with the displacement of blue-collar workers. However, it has become evident that a substantial wave of automation is approaching various industries, including white-collar professions in areas such as law, finance, software development, tax accounting, and many more. As a result, some roles may undergo significant transformations, while others may even become obsolete.

Companies that proactively acknowledge this trend and invest in automation technology early will likely gain a competitive edge, capturing a significant portion of the market. Conversely, those who fail to adapt may face the risk of losing out. It's crucial for organisations to recognise the potential impact of automation on their operations and strategise accordingly to navigate this changing landscape successfully.

Share this post
Curvestone

Similar articles

Accelerate your document analysis like never before