Generative AI – Creativity vs utility… and contact centres (of course)
I can say it, I’m getting older (perhaps wiser… some may disagree), and the whole AI term is now thrown around as part of modern culture and acceptance, some don’t understand it and some don’t want to know it, but one thing I am seeing is how it can be useful in my line of work.
If anyone remembers AI from the 80s, Bladerunner, Terminator, and Wargames usually spring to my mind, still to this day it manages to set a scene of what we ‘thought’ AI would be when it grew into our century, although farfetched, some of it remained, it was a balance between creativity (out the box thinking) and utility (what it was used for helpful scenarios), usually most of these films saw it as a threat of some kind but many didn’t anticipate how it would be used today for the positive side of life in assisting in every area (the actual utility).
I ramble a bit when it comes to this, but it’s a really interesting topic when applying this technology to my consultative role when speaking with clients about it, there doesn’t seem to be one right ‘answer’.
We can see most vendors on the market are either adapting towards the area or are already embedded alongside it, but the true reason for this is that it’s a trend, and in most businesses these days, if you’re not on the trend line you’re falling behind or off it completely targeting another niche in the market commercially.
The history of Generative AI has gone through a lot of changes and challenges in the past, having a good read up on it over the past year, for me, has allowed me to see, in my view, how and why it adapted towards it becoming usable in everyday tasks.
This goes all the way back to the 50s when it was first used to play checkers (a game, I know…) and the term Machine Learning came into the ether of words and is still around today, I find it fascinating that its purpose back then, to work out the best next ‘move’ still applies today in the Contact Center field.
A few things come to mind when I think about the application on where and how this technology can be used, and of course, the Dynamics 365 movement by Microsoft has shoved this into the next generation with ease of access, some other vendors that I’m close with are also bringing it inside of their products (Anywhere365 and Luware) and it’s great to see it is moving with the requirement factor… but more on this one below…
One of my biggest gripes I see currently is that the technology itself, from a creative perspective, is far ahead of the utility possibilities, what I mean by this is that where it’s trending and everyone and their dog is using it, it’s taken a leap from not understanding what we can actually ‘do’ with it from a practicality sense – hit and hope.
That takes me to use cases…
Being a consultant for coming on 18 years now (that long…) has really helped me adjust how I like to engage with new and current clients with the CCaaS market and Generative AI overall, one of the big questions I always like to go in with off the bat is “what are you trying to achieve with this…”, it throws many people back as frankly, they don’t know, what this does is allow discussion, insights, and roundtables to get the business process’s flowing – this is the part of consulting I love, it enables all individuals in any part of the business to have a voice.
I firmly believe that for value to be had with any of the AI tech that’s hitting, it needs to provide value, and that value really comes from the individuals seeing the outcome to make a difference… next time you’re sat next to one of your fellow agents, or as a supervisor get them around a table, just ask… how many of the same queries do we keep getting through that can be answered with an automated response, I can guarantee many will reel off a list but most importantly this will allow them to focus on the other queries that get taken over by the repetitive – now this, is utility (value).
Moving onto the other side of Gen AI, the creativity, the tone of a response, or the general feel around how it deals with questions is huge, the technology can grab onto a bunch of data and come up with a possible resolution… sounds great right?
A couple of niggles that come out of the other side of this way of working for me, one thing I notice a lot when talking with clients is that internally most of their data is a mess, not just from a content perspective but organisational as well, what this does is give us some approaches which sometimes don’t work out:
- “Stick and See” – This approach makes it easy to have a ‘throw’ approach when it comes to letting GenAI read up on this information, in doing this can possibly compromise the data, have no control over the output, and can be false leading when it tries to cobble up an answer, instantly killing the use case for using it and already leading onto the back foot.
- “The Trust factor” – When the solution comes into it, it will try and set a tone (which we can control too) with the output, statistics say you have a very limited time to have a potential new customer or query answered before it comes through to an agent or they start to lose confidence in its ability to get an answer, when it either appears too happy, spouting off nonsense which is irrelevant to the query or is just too ‘over helpful’ we enter the cringe phase, a term I love to use someone when loses the trust quickly, this in turn just defeated the object of why GenAI should be used, to deflect away from the agents again.
So why does Creativity pose such a possible problem in this field, well, it comes down to everything I started out on, the use cases, remembering those things I said what the goal is of using this tech, to make every individual’s life easier… trust is a huge factor when it comes down to it, creative it not always warranted when a goal needs to be met – present the facts without the cringe!
In my part of the field with Contact Centers, all of the above is based on my own experience over the years, I’ve seen systems go from 30 nested deep IVR solutions to no IVRs at all, voice-activated unified routing to long queue wait times with skills being understaffed, but the questions I always stray back too is – you guessed it – use cases, understanding why you want to use this is the key.
It’s why Microsoft has played a blinder when it comes to Dynamics 365 in general, not just from the new contact centre offering but the whole Dynamics Applicational standpoint, it’s worth checking out their separate targeted apps such as Dynamics 365 Sales, Finance, Marketing, Field Service – all the underlying technology is slowly being introduced into these applications because it serves a ‘need’ for a particular field, they did their research and targeted the markets for those use-cases (7th time in this blog saying it), all the new tech may not be in each one yet, but give it time….
The other vendors I am close to such as Anywhere365 and Luware all pitch into these new NLP platforms as well, it’s great to see how they can be pushed down to an agent or help deflect a customer away, deflecting is not a bad thing, it’s a misunderstood term, it’s saying “We want both sides of a call to get resolution faster” – nothing more, nothing less – sounds productive too me!
I’m always up for a good discussion on the subject from all and any that have an interest in where it can go or how it could apply to your daily lives around the Contact Center area, look me up on LinkedIn (icon at the bottom of the blog if you fancy a natter!
Until my next rant, and more memes, bye for now!
Ben