Viewing entries tagged
AI in communications and PR

Comment

calling centaurs & cyborgs: time to recognise ai collaboration in comms

When it comes to AI, there are no shortages of hot takes on what it all means for society - for better and for worse.

by Alex Waddington

Our jobs, our health, even our relationships – they are all in scope for impact and disruption by Generative AI, if the experts and headlines are to be believed.

The International Monetary Fund (IMF) has warned AI could affect nearly 40% of jobs, replacing some and changing others. OpenAI founder Sam Altman has even been quoted as saying that AI is going to replace "95% of what marketers use agencies, strategists, and creative professionals for today". Gulp.

But when I looked for blogs on AI on comms2point0 there were only two – one from myself, and one from Ian Curwen. Considering it’s nearly three years since ChatGPT launched and changed everything, that’s a little surprising.

Does this mean public sector organisations are behind the curve on AI adoption? Are comms professionals scared of it? Are they being frustrated by internal processes and risk aversion? Are people using it regularly - but unofficially and very quietly?

The answer is, I suspect, yes to all of the above.

Sorting substance from sensation

Part of the challenge is the sheer noise about AI and what to believe. Last year PR legend Stephen Waddington of Wadds Inc (no relation) called peak hype on AI. Almost a year on, that hype seems alive and kicking, with AI being deployed with gusto into more products and services we all use every day.

And only the other week, UK Technology Secretary Peter Kyle breathlessly declared that the UK “can entirely rethink and reshape how public services help people through crucial life moments using the power of emerging AI technology… we could provide a level of service to citizens across the country that was previously unimaginable”.

Thankfully, we are seeing emerging use cases and sector responses that can help us sort the substance from some of the sensationalism.

Showing what is possible

At a national level, the Government Communications Service (GCS) has been leading the charge for AI in public sector and not for profit comms, showing what is possible.

In May, it revealed its Assist tool (based on Anthropic’s Claude) was being used by 4,500 users across more than 210 government organisations - over 70% of all potential GCS users. What is more, it says the average comms professional in the civil service is saving around three hours per week by using Assist.

Projected annual efficiency savings for GCS users are £5.5 million. Imagine what you could do and achieve with just a fraction of that cash saved.

Use cases for Assist include developing draft crisis communications plans, conducting COM-B analysis for behaviour change campaigns, and testing material for accessibility. I met one former GCS member recently who admitted they were struggling to operate without it in their new role, such were the benefits.

In the NHS – where we’ve seen communications roles in the political firing line in recent months – a Communications AI Network has just been launched. There is also the development of a national operating framework, an ethics framework and online training hub.

A report published to mark the launch of the new Network noted AI is being used to simplify technical language and analyse feedback from patients and staff. It also found widespread informal use of AI tools such as ChatGPT outside of formal approval processes and governance.

Growing curiosity and exploration

In the challenging and budget-constrained world of not-for-profit and public sector comms, I detect a growing uptake, curiosity and enthusiasm around Generative AI and its many possibilities. Not to replace what we do, but to enhance and complement it.

Every workshop I run about using data in comms, the AI part tends to generate the most discussion and curiosity. Webinars on AI organised by LGComms earlier this year were in high demand.

From conversations with communications professionals, Copilot is starting to be trialled and rolled out as part of the Microsoft corporate workplace package.

Comms teams are also putting their hands up to be involved with policy development, piloting and roll-out – which they should be, of course, given the reputational risk of botched deployment or starry-eyed adoption in blind faith.

And so, with these visible developments and examples of innovation, it feels like the right time for a new ‘Best Human x AI Collaboration’ UnAward, proudly supported by Whetstone Communications.

Yes, there are most certainly ethical, legal and environmental considerations you must deal with before letting Generative AI loose on real work and projects.

But as GCS and now others are showing, with the right guardrails, tools and training in place, many things are possible. You could free up time for more interesting work that draws on your human strengths, and use it to unlock fresh ideas and creativity, even at times when you are not at your most sparkling.

And if you are only just starting your AI journey now, you can benefit from the work and lessons already learned by the early adopters.

Centaur or cyborg?

In his excellent book Co-Intelligence : Living and Working with AI, Prof Ethan Mollick talks about centaur and cyborg approaches as being key to gaining the benefits of AI without the disadvantages.

Centaur work has a clear line between person and machine. Centaurs have clear red lines on tasks they will allow AI to handle, and tasks that must be done by humans. An example might be the creation of a campaign plan to increase recycling, where you give a very detailed prompt with clear parameters of what is needed, based on your very human lived experience, knowledge and insights. The AI quickly produces a plan (with helpful suggestions), which you check, seek further feedback on, and sign off in the usual way.

On the other hand, cyborgs blend machine and person, integrating the two deeply. Cyborgs don't just delegate tasks; they fluidly and rapidly intertwine their efforts with AI, as a successful partnership.

Cyborgs find themselves working in tandem with the AI depending on the strengths and limitations they have identified – and of course, as technology develops (the release of ChatGPT 5, for example) the balance and nature of this may change.

An example of cyborg work might be the creation of a campaign plan to increase recycling rates, where you would entrust an AI to do ‘deep research’ and use academic literature on what works. You would then challenge and prompt in turn, iteratively refining the plan by combining the AI’s speed and powers of reference with your own critical thinking, judgement and practical considerations like budget.

Will AI come to the UnAwards party?

To enter for this new award, you don’t need to have created your own version of GCS Assist and saved millions of pounds – although hard examples of real, tangible impact will certainly impress the judges.

Examples of successful AI x Human collaboration might be in supercharging a successful campaign with initial AI-powered research into ‘what works’ (and importantly ‘what doesn’t’) for creating behaviour change among hard-to-reach audiences.

You might have used your corporate instance of Copilot to overcome a lack of in-team data literacy, to crunch large public datasets and find important sparks of insight for effective local audience targeting. Or maybe you used the tool to analyse several hundred open text survey comments, to find hidden themes, shaping new approaches to internal communications and employee engagement.

You might even have deployed AI in your planning and stress-testing crisis scenarios, using it to generate unforeseen events that really put your team through its paces. AI might have helped identify training needs, ultimately resulting in your team being more prepared and resilient.

Or perhaps you’ve helped more of your residents or customers get the support they need by creating a chatbot that has breathed new life into hundreds of pages of dense but valuable content sitting online.

When all said and done, the UnAward for Best Human x AI Collaboration hopefully does three things;

1)  shows good ways forward for safe and ethical AI-human collaboration in public sector comms

2)  encourages teams and individuals to be bolder and more positive, proactive and transparent about their use of AI, and

3)  showcases to sceptical stakeholders the outcomes and benefits that can be achieved when the technology is made available to skilled communicators

Ethan Mollick talks about the core principle of always inviting AI to the party - even if it ultimately only plays a minor role, or you ultimately decide to discount what it suggests. Only through this approach, can you really learn about what the technology can and can’t do (yet).

When the UnAwards winners are announced later this year at the famous and fabulous winners party in Brum, it’ll be fascinating to see whether AI is up there in the limelight taking some of the plaudits - or still stubbornly lurking in the corner.

Alex Waddington is owner of Whetstone Communications, a sponsor of the UnAwards 2025

*Sign up for the comms2point0 eMag*

The comms2point0 eMag features exclusive new content, free give-aways, special offers, first dibs on new events and much, much more.

Sound good? Join over 4k other comms people who have subscribed. You can sign up to it right here

Comment