AI |
||||
|
3.5 model. I still remember the‘ wow’ moment when I pasted one of our text blocks into GPT 3.5 and received a concise company summary in return. It felt like discovering a secret door in my own workshop; suddenly the business had a scalable way of breaking shows into their component parts and letting data define clusters. This was a massive breakthrough.
Scaling processes with an army of AI interns One of the biggest mindset shifts was the realisation that we had an infinite army of AI interns to use on simple tasks. If we could break work into small enough units, there was no limit to the effort we could apply. Much of our learning over the last three years has involved working with these AI interns. Realising how they misunderstand our prompts, how they behave when asked the same question many times, and making subtle adjustments to improve outputs. The payback has been massive: our prompts and processes now tackle thousands of repetitive tasks each day, from keeping records up to date, creating new data and refining existing data to extracting fresh insights. While early results were often frustrating, continual experimentation, combined with ever improving models, has built something genuinely impactful.
Over the last year, I’ ve also found that, alongside an unlimited pool of somewhat obtuse interns, our business has benefitted from code developed with a skilled AI co-pilot programmer. While our core database and infrastructure were built by someone far more technically gifted, AI has allowed us to‘ vibe’ code using prompts. This has produced countless new routines on top of our data without the need for further technical assistance.
We’ ve relied on AI to write thousands of lines of Python from a simple description of the need. These are small code snippets that transform, embed, refine, cluster and describe data tirelessly. This co creation of
|
code is immensely powerful because the only limitation is the clarity with which you can describe your aim. By layering concepts, AI coding can add complexity until you achieve your outcome. Sometimes it’ s slowgoing, but by turning many manual processes into code, we’ ve eliminated hours of drudgery and achieved far deeper insight.
People + machines: shaping the exhibitions of tomorrow AI hasn’ t just liberated the data refining and project delivery day job, it has reshaped our sales and marketing approaches, too. While I still shy away from AI generated posts( content often bland, hyperbolic and off voice), as an idea generator and challenge partner, AI tools now excel. Discussing business problems with AI has led to smarter CRM deployment, tailored marketing collateral and a prospect scoring system that simply works.
Perhaps most transformative has been the way AI has forced me to rethink hiring. In the past, I’ d often recruit before fully defining a role. Now, I sketch processes first, decide where AI can handle the routine tasks, and then hire for the parts that demand process management, empathy, creativity or judgement. The result? A leaner, more flexible( if smaller) team working hand in hand with our digital workforce.
The pace of change has been staggering. Over the past year we’ ve added more exhibitors than in all our eight previous years combined, created 250,000 + AI generated exhibitor biographies and rolled out new better clustering approaches. From a business perspective, blending code and automation allows us to deliver HuntEx projects concurrently.
For exhibition organisers, I hope our journey resonates. The message is simple: examine your processes and experiment to find where AI can drive scale or insight. Accept that AI R & D has a cost and will sometimes fail, because
|
“ It’ s about the late nights, the light bulb moments and the frustrations that come from weaving AI, prompts, code and agents into our processes, a journey that, I suspect, many organisers are also making right now” |
when you find tasks your AI interns or co workers can deal with, their impact is game-changing.
Looking ahead, the obvious question for exhibition organisers is what remains after deploying AI. This is a topic I’ m eager to moderate at the upcoming EN Indy Summit on 17 October in London. The question we’ ll explore with a panel of experts is:“ Which distinctly human skills will remain our greatest asset in an AI driven world?”. My initial perspective is that the ability to design business processes, intelligently carve out and delegate tasks to AI, and invest in those areas where the human edge makes a difference is key. If my journey has taught me anything, it’ s that the future of building data led events belongs to those who master both technology and humanity. EW
|