Generative AI (GenAI) takes traditional automation applications to the next level. Traditional automation applications are rule-based. You define the rules and actions, then configure the software to follow them. The structured sandbox limits flexibility in handling exceptions. In many cases, an exception report or alert is used to manage uncommon scenarios that require staff time to address.
With GenAI tools, they are trained to handle different scenarios, including exceptions, using patterns learned from vast datasets. This reduces human intervention. However, The output could be “creative” and not necessarily best aligned with your brand.
As a result, skills for working with GenAI are different from traditional automation technology. There several essential skills.
- Prompt Engineering. Good output requires thoughtful prompting. A generic prompt gets you a generic output. A good prompt generates an output that is applicable to the scenario you have on hand. There are many parameters to consider. Some of the common ones include role of the user, context of use, tone, audience for the output, level of communication, level of details, format, and brand voice. Quality of output increases with specificity.
Traditional automation applications require specific input data to drive the pre-defined actions. The data fields are set up so users don’t need to consider additional input parameters in order to move to the next step.
- Output Evaluation. GenAI output varies with each prompt. Users get a different output even when they use the same prompt. Subject to the datasets used for training the AI model, output could be bias. Therefore, you need to evaluate the output for bias, relevance and appropriateness. The subjective evaluation aspect requires individuals who have an aptitude for it. Using GenAI output without proper evaluation and fact-checking could be disastrous. Hallucination is a constant risk.
Output from traditional automation tools are defined by rules and logic. The output is predictable. You challenge the output when you feel the rules and logic are no longer valid.
- Problem-solving Approach. Abstract business challenges could not be addressed by the traditional automation tools because they are hard to define in a systematic manner. With GenAI, you can prompt for creative ideas by asking open-ended questions. In fact, you can set the temperature and let the model be as creative as it can be. GenAI is great for generating ideas. Hence, you need to be cognizant when GenAI is suitable for the use case.
A well-defined and engineered automation application takes care of the tasks it is built for. Your staff handle exceptions when the application detects them.
- Human-AI collaboration. With GenAI output, you review it and prompt for refinements. It is an iterative process. It is possible that you get overwhelmed by all the ideas. It is important to be able to select the best idea(s) and move forward with next steps. Otherwise, you could be spending hours chasing the “perfect” idea. The skill on orchestrating information and take action becomes important.
Traditional automation applications are designed to execute tasks following the set rules and conditions. There is little need to collaborate other than changing the configuration when necessary.
- Adaptive learning. The advance of GenAI technology is going at such a fast pace that new capabilities are launched every day. So keeping up-to-date on what your business could deploy to build competitive advantage is important. That said, you need to have a strategy on how to leverage GenAI technology early. Continuous up-leveling skills and knowledge would help to overcome FOMO (fear of missing out) and be selective with what works best for your business.
Traditional automation software vendors are actively incorporating AI features into their products. This takes some pressure off your shoulder but it is still important to keep abreast of what is possible with GenAI.
GenAI has introduced new dimensions to consider and tackle business problems more proficiently. The flexibility in handling text, image, audio, video, and actions boosts your ability to optimize how to use your data across platforms. You need more than good analytics skills but a broader set of skills to take on the opportunities GenAI offers.
If you find this useful, you might like 5 Skill Shifts for Digital Transformation.
Leave a Reply