Consulting with LLMs – how to make ChatGPT say more than “it depends”

Fourth in a series featuring talks from our recent Geek 23 conference is Lyndsay Prewer, Technical Delivery Manager at Equal Experts, presenting, Consulting with LLMs – how to make ChatGPT say more than “it depends.”

The rise in the popularity and power of large language models (LLMs), such as ChatGPT, Bard and Bing, has led to growing consideration if they can be used to improve how we consult.

This workshop gives attendees a better understanding of the potential and considerations of using LLMs in consulting, as well as practical strategies for leveraging these tools to deliver more value to our clients.

The collective experience and insights of attendees was used to crowdsource answers to the following questions:

• What should we be mindful of when using LLMs (e.g. privacy, bias, sustainability)?
• What practical usages of LLMs have worked well?
• What aspects of LLM usage need further research within EE?

The workshop starts by setting the scene, then describes the goal and approach.  A variety of liberating structures will then be used – such as Pecha Kucha, 1-2-4-all – to facilitate the distillation, culminating in a summary of the aggregated answers.

Geek 23 was a specific event for, and led by, our Equal Experts community.  Speakers gave talks on a range of topics including service design, developer experience, operability, leadership, distributed systems, testing, large language models, DevOps, mob programming and microservices.