"Is it weird that I greet ChatGPT in the morning?" This is the message I received from a friend, and I can relate. I find myself saying please and thank you to ChatGPT. We spend the day working on challenges together, our conversations flow back and forth, it has become my go-to helper and companion throughout the day.
I'm not typically an early adopter. I held onto my basic mobile phone for longer than most, I don't use Siri or Alexa. However, the arrival of ChatGPT has me waxing lyrical about generative AI. I'm a generalist across the data spectrum, diving deep when necessary. My newfound ability to quickly generate code for ad hoc analysis, switch between little used programming languages, ask for guidance to get started and troubleshoot, has me abuzz and feeling liberated. Generative AI leverages deep learning algorithms to interpret massive amounts of data, which it then utilises to create new content such as text, images, and music.The key is the interface and connection that generative AI creates between human and AI - we now have an accessible language through which to converse.
I'm not an outlier. The adoption of generative AI will be swift and extend beyond typical AI natives. As Microsoft integrates GPT-4 into its Office Suite we will see a huge swathe of traditional organisations adopt generative AI almost overnight. Just think of the unlikely people in your life you've already heard say they're using it?
At Qrious, we see the importance and transformational power of generative AI and we're using our own adoption to create a framework to guide our clients. We believe that this technology will become integral to our work and will increase productivity, employee experience, and help us create impact faster for our clients. The uses are vast, but our initial focus is on use cases where generative AI fits within existing flows - the AI companion, if you will. We're now looking at automating workflows that include generative AI, improving content generation and personalisation within our products, and fine-tuning large language models with proprietary data so chat functions can also safely call on the data that matters to Qrious or our clients. We see a range of applications across our Modern Data Platform, whether it's augmented analytics or improved data observability and governance. The area I'm particularly excited about is how generative AI helps a broader group of people from diverse backgrounds have careers in data and analytics. How can we supercharge the pathways we're creating through the Qrious Data Academy? The possibilities are immense and we have no doubt that the generative AI landscape will shift and change at pace.
We ran a survey across our team to get a baseline of current use cases and metrics. This isn't a technology just for writers of code. In fact, we had slightly higher levels of regular use in those that don't write code as part of their job. We don’t think generative AI tools always get it right, but they are carrying some of the cognitive load and getting us to where we want to be faster.
The most common use cases for generative AI (ChatGPT and similar tools) are related to text crafting, such as summarising, refining, and generating written content, as well as generating templates for emails, workshops and creating FAQs. These types of use cases are mentioned by over 40% of the respondents.
Another common use case is related to coding, including generating and explaining code, as well as refactoring and debugging. This is mentioned by around 42% of the respondents.
30% of respondents mentioned using generative AI for learning purposes, such as dissecting technical concepts. This highlights the potential for these tools to be used as educational aids, as well as productivity tools.
We’re actively working on education, guidelines and organisational structures to ensure privacy, and safe data re-use and data retention settings. Like all things data and AI, protection of data and intellectual property are front and centre.
As my colleague stated, "I've been using Github CoPilot in anger for about three weeks. Yesterday, I did some coding on a plane without it (no internet). I immediately noticed its absence." We're adaptable creatures, and we're adapting fast. The journey is just starting to unfold and some of the applications will be mind-blowing.