The UK union movement recently recently saw a huge and incredibly important event – the launch of the Employment Rights Bill by the new Labour government, which promises to expand a number of important worker rights.
At the Unions21 Trade Union Renewal Conference this week I presented a series of examples of how different capabilities or features of one generative AI tool (ChatGPT) could have been used by different departments as a partner to support the kinds of work they might have to undertake in response to the Bill.
The purpose of these demonstrations was to find use cases that the union staff in the room could connect with: things they might also want to do.
As I explained to those in the room, understanding at an individual level what generative AI can be used for matters. Generative AI is highly individualistic: it can be used different tasks for different people, and workers most like to use it on the tasks where they derive no meaning... and which tasks give meaning (or not) differs from person to person.
This is something I believe we as unions need to understand for ourselves: and seeing (or using) is believing.
“I think unions will need to actually use AI for themselves to be stronger in the AI age… the unions who know best how to effectively and ethically use generative AI internally may also be best placed to organise and recruit those many, many workers most impacted by generative AI in their workplaces. Workers who are among the least unionised at present.
1. Policy: summarising and creating a spreadsheet for team (using ChatGPT code interpreter)
The first task was among the most obvious, and was directly inspired by conversations with union staff about the work they’d done in response to the Bill’s release: in Policy departments they’d had to start planning their response. The important thing here was that we weren’t replacing the reading of the Bill, merely replacing a small task in the process: creating a first draft spreadsheet to assign team members too.
I then showed four other areas and tasks, the videos of which are included below.
The team administrator creating an interactive and innovative team meeting agenda to share viewpoints and responses to the Bill.
The communications officer identifying a range of different options for content to be written to engage members around the Bill.
The organiser needing on-the-go coaching and support on handling a difficult conversation with an activist unhappy about the Bill.
The strategy lead thinking about how the Bill could be used to further their goal of digital transformation in their Union (this one’s very close to my heart!)
2. Admin: planning an interactive, innovative team meeting (using a GPT)
In this second example you see the capability of accessing specialist knowledge outside a staff member’s usual areas of skills and knowledge. In this case, facilitation methods and techniques, loaded into a “GPT” - which is a version of ChatGPT tailored with specific data for a specific usage.
3. Comms: ideation and content writing (using ChatGPT with Canvas)
Content creation is perhaps one of the most well-known uses of tools like ChatGPT. However I’d say that it is often the worst usage: the text rarely is as good as a human, unless great care is taken. What can be really useful, however, is inspiring humans with different examples of different ways to write a piece of content. That’s what is shown here - along with a quick introduction to one of the latest ChatGPT paid features, the “Canvas” where you work jointly on a piece of text.
4. Organising: coaching for member conversations (using ChatGPT Advanced Voice Mode)
Many union staff may have moments where they need basic advice and support and have no one to talk to. For example an area organiser driving around in a car on visits to branches. Here’s an example of where generative AI can fill that gap, and offer advice that is useful because it covers a very common challenge: how to handle difficult conversations.
5. Strategy: digital transformation programme impacts (using o4 advanced reasoning)
Finally, using the ability to AI to work through more complex challenges could be useful in terms of thinking through different approaches to strategy. Of all the demonstrations on this page, this is probably the least successful, as it uses a ChatGPT model (o4) that only has data to October 2023 and doesn’t allow for PDFs to be uploaded. It therefore doesn’t actually have the data to know what was in the Employment Rights Bill - and not. However you get a sense of how a person in a role could use generative AI to quickly think through how a big event, like the Bill’s release, relates to their strategy in an area that is relatively far removed from it – like the digital transformation programme of a union.
