Wouldn’t we all love a personal assistant? Someone to sort through our emails, to manage our appointments and to remind us of all the important things we need to do. Some fortunate ones can afford assistants, but most of us simply have to go without. Unless… what if a large language model (‘LLM’) like SCOTi® AI could take over the role of being your very own personal assistant?
LLMs could remove the need to juggle multiple sources and programs used for organization and instead provide a streamlined service. They could do everything from prioritizing your emails, to providing data analysis in plain English, and even help with generating reports. The benefits of using a large language model are that unlike a human virtual assistant, LLMs are available 24/7, highly scalable and provide an immediate response. However, they do struggle in some tasks that require a more human touch such as complex customer service issues or creative tasks like graphic design. Companies should consider what qualities they would want to prioritize in an assistant or even consider using a mix of both.
Examples of effective use of LLM as virtual assistants include a marketing assistant to create social media posts, or a software developer to provide code snippets. It should be noted, however, that the utility of these assistants is dependent on the way in which they are used and the quality of the prompts used.
While some hail LLMs as the development of the century, others are disappointed and unimpressed by their functions. It is certainly true that the technology behind LLMs is still being developed and there are certainly limits to their use, but these constraints are also defined by the skills of those using the technology. Like all tools, LLM’s full potential can only be obtained by using it for the purpose in which it excels and through some adeptness in creating the best prompts. We would be lying to ourselves if we pretended that LLMs had attained a mastery of natural language so great that no skill was required in creating prompts.
Tailoring good prompts and being able to effectively use tools like SCOTi is such a high valued skill that a new market of AI prompt engineering has been created. The ability to effectively use these tools is considered so desirable that some warn that not possessing this skill when colleagues or competitors do may be the real threat behind AI. It is not farfetched to imagine a future where skills in effectively using generative AI chatbots are as standard as the ability to navigate a word document.
A collated list of tips and tricks for improving prompt creating skills can be found below:
Check out the links from the tables:
- General Tips:
- Use buzzwords
- Specifications to narrow down your results
- Questions VS Statement
- Useful Prompt Ideas
Written by Celene Sandiford, smartR AI