Simon Clayton is not convincedArtificial intelligence (AI) has been the stuff of sci-fi for decades and is now finally making some interesting leaps into the consumer world. Seth Shostak, Director of the Search for Extra-terrestrial Intelligence Institute (SETI), says: ‘Within 20 years, you will have one computer that's smarter than all humans put together.’
AI is used to solve complex problems where there is no set pattern (analysing handwriting for example), so the computer has to ‘learn’ how to recognise a trait and respond accordingly.
The term ‘AI’ is, however, often misused, just like the term Big Data – but that’s a topic for another column. Real AI is where a computer is constantly fed data that enables it to process the outcome of an action and where it then applies the outcome to future actions - thus it ‘learns’ from each action that it processes and it adapts each time. You have to keep feeding it data, so that it can continually learn and ‘evolve’.
If a programme creates variable outcomes based on variable answers - for example ‘if X equals 3 then do this, but if X equals 4 then do this’ - it is not AI. That is just standard programming. And there is a massive difference. Amazon does this well: when you buy something, it will helpfully suggest other things you may like to buy to complement your purchase, such as a hardware product and the batteries it needs, or linking one film to another of the same genre.
But AI has now reared its head in the events industry. According to the press, the world’s first AI event networking solution has been launched, which uses AI to match one visitor to another visitor or to an exhibitor at an event, so they can network.
For AI to work in this situation, the programme would have to be fed a constant stream of data – not just a one-off hit from a registration form. Visitors would have to collaborate and keep updating the app with feedback, saying whether or not they found the suggested meeting helpful.
The fact is that events are ephemeral. Most change from year to year . Certainly around 30% of the exhibitors change. The result is that what is learned from one year couldn’t be applied to the next. Nor does the process take into account contextual data or personal preferences or experiences.
I’m now going to sound like a Luddite, but even if this could happen, isn’t this a step backwards? When hosted buyer events first became popular, exhibition organisers would match hosted buyers to their exhibitors and tell them which companies they should see – all based on their understanding of a visitor’s needs from the forms that had been filled in.
Being told who to go and see didn’t really go down that well with visitors and some very senior people in the industry complained (rather publicly) about this antiquated practice. They stated, quite rightly, that they were grown up enough to decide for themselves whom they really wanted to spend their time with. Some industry people even said that they would prefer to fund their own trips rather than be hosted and then forced to see companies in which they had no interest. That matching system just didn’t work.
IMEX is the largest events industry exhibition in the world. Their hosted buyer diary system was created for them by RefTech more than ten years ago and it is the backbone of their two annual events, IMEX America and IMEX Frankfurt. It allows buyers to review exhibiting companies and then to choose whom they want to spend their time with. IMEX trusts buyers to be in charge of their appointments, to know their event objectives and then let them decide who can help them.
IMEX buyers use the website’s search engine to search for exhibitors under a category (‘Conference Venues’, for example) and then manually go through that list, looking at their images, reading their information and so deciding whom to meet.
Using AI to take this decision out of an educated person’s hands is ludicrous and a great example of using tech for tech’s sake. Or using a sledge hammer to crack a million nuts – most of which you don’t need. I’m pretty sure it’s not real AI and they are just using exaggerated tech speak, but I would be happy to find out more and be proved wrong. If the developers are reading this, then please get in touch, because I’d love to see a demo of your AI. Either way, it is going on my ‘Ones-to-watch’ list: in a year’s time I will review it to see if it has actually made any impact on the industry and the way we network at events.
But I’m not feeling very optimistic.