Webmart has never been shy about adopting new technology, nor in trusting youth. The Bicester marketing business has a team of programmers, customer account handlers, designers and project managers who are all young and are presided over by the not so young executive chairman Simon Biltcliffe.
Although the business was launched in the pre internet era, it has successfully managed to ride the waves of successive technologies like Donkey Kong leaping from one swinging platform to the next.
Now Biltcliffe is ready to embrace the coming artificial intelligence tsunami. His role keeps one foot in the business and the industry while he is able to place another foot in whatever passing pie looks intriguing and interesting enough to feed back into Webmart. Artificial intelligence is one of those pies.
He uses ChatGPT, the best known of the large language model AIs, to summarise blog texts and to translate to and from English to around 50 other languages in real time. “It does this really well,” he says.
“We are going to see AI used for automating tasks, for data analysis, for improving decision making and for making your organisation a lot more efficient than it would otherwise be. In short it is a really good way to go from gross profit to operating profit.”
He believes that Webmart will embed AI into how the business operates over the next few years. It will be about automating back room tasks, those that are not customer facing facing, and so freeing up resource to be creative, to communicate and to deliver a better experience for customers.
Staff are already being trained to think how AI can change their roles and in turn this will aid recruitment of the next generation who know they will be joining a business that is constantly pushing ahead. It also gives Webmart a chance to recruit from a pool that might otherwise want to join Google, Amazon and the other tech giants. “You need to have intelligent people who know how to query the AI and can then validate the output they receive,” Biltcliffe explains.
Currently in some areas, academic research and law in particular, the LLM AIs have generated papers and legal arguments that reference non existent academic papers or cite precedents from court cases that simply have never taken place.
These are called hallucinations and while they will eventually disappear, they can currently trip up the unwary. When using AI to write software code, something where the technology can save hours, days and weeks of programming and checking the code, someone will need to ensure that the code is good – and for the most part it is.
For Webmart this will make it easier to add new features to the benefit of its customers through the speed of code generation. If it did not, by implication others will. For Biltcliffe, using AI is essential to the business. “We need to be an organisation that embraces AI. It is another inflexion point like the arrival of the internet, only AI is much more powerful as a disruptor than the internet has been.”
Tasks like ensuring a job is to the correct specification, that all elements are handled correctly, that the job is sent to print on time and that it is then produced and delivered on time, with the timings and costings and now the environmental impact, measured at every point, will be part of the AI set up. Webmart has, over its 30-year history, always gathered this information and used it to make sure that jobs are placed with the most suitable printer for that job. Its success has been down to the accuracy of these calculations and its ability to take into account factors that might change quickly.
At one time this was data about every web offset press in Europe, its cut off formats, its speed, location and so on. There would always be room for error, a press offline or fully loaded at the wrong time for example due to the latency of communication by phone or email. AI promises to wipe that away through machine to machine communication, finding a snugger fit between job and producer than ever before.
But it is not an AI only future. AI, says Biltcliffe, works on this type of task but cannot do everything. “There’s no need for a human if there’s no compassion needed,” he says. “Where there’s a need for creativity, there will be a blended approach. Where the need is for compassion and empathy, the human is needed.”
But this human, the one that listens well, that displays the soft skills (“where we have never really invested enough”), will be necessary to plant the right information in the AI. This is the reengineering that is underway for Webmart. The company is aiming to reequip its people with the skills that are needed to thrive in the AI age. “It is much better to repurpose your team to be fit for the AI world than to sack them because they lack the relevant skills. If you understand that the only unique value that humans will bring are around compassion and creativity, you have to train for that and train them to be empowered by AI, not replaced by AI.”
Leading in this way will bring a business benefit as Webmart will be positioned to help its customers cope and transition to these sunlit uplands, because all businesses will be affected in some way. More often than not, they will be less agile than Webmart, and can take direction from what it has learned.
Biltcliffe has himself been on a knowledge pilgrimage to understand the implications of AI, talking to university professors, networking with other entrepreneurs, reading and watching widely. This has been distilled into the course of action for the business.
“There are four areas where AI can have an impact on a business,” he says. “That is in customer service, sales and marketing, software engineering and research and development. Printers frequently do the first and second, then the third and perhaps the fourth of these.
“For customer service, AI brings real time data to job management, to job submission and so on, but when you really want to understand the underlying strategy driving your client’s decisions, that’s when you need humans.”
If, however, AI promises transformational change, there are challenges on route. Data may not always be in suitable condition for sharing across systems. There needs to be an agreed standardised approach, a process control approach within agreed guidelines. “A huge challenge the industry faces is a sheer lack of talent, another is budgetary constraint.
“Fortunately,” Biltcliffe continues, “we are not creating AI, we are using it.” That will ease the need to recruit the very best computing graduates, though print will still need to show potential recruits that it is an industry with a future.
There are other risks that need to be understood: data and cyber security will need to be water tight because some will be using AI to penetrate firewalls and the like with ever more sophisticated attacks.
There’s a danger too that the use of technology can lead to an over reliance on it, opening up pitfalls like the risk of hallucinations when AI gets it wrong, like drivers dumbly following their satnav as a country line becomes a muddy track and eventually an even muddier field.
MORE FROM THE MAGAZINE
MORE ON WEBMART
