Breaking Barriers - Embracing AI and Automation for Business Transformation

Joining Tom Henson, Managing Director of Emerge Digital, and Peter Quintana, Director of hgkc, were:

• Mark Allen and Paul Hague from garden furniture sellers Bramblecrest
• Rhoda Brain from Bristol-based digital agency Miint Marketing

• Matt Hamilton and John Lee from home automation experts AutoAbode
• James Taylor from Geotechnical, experts in ground investigation
• Peter Marchbank from rotary precision instrument engineers RPI
• Steve Simpson from sub-sea and rail engineering specialists Viper Innovations
• Julian Lee from hydraulic solution engineers DCA, part of the Fluid Power Partnership

• Mary Lee, partner at Oslers Solicitors
• Ed Surman from business services consultants Mushroombiz and Tlam
• Adam Starkey, owner of food innovation company Green Gourmet
• Jess Brewster, representing Roses Theatre and the Tewkesbury Cultural Consortium
• Kim Jones, co-founder and innovation expert from management consultancy hgkc

Some attendees said they were there to learn more about AI, some had been using it for a while, and some were wary of the potential threat presented by the rapid advances being reported.

Rhoda Brain is already using AI in their operations and said it has greatly impacted their marketing service offering, adding that their firm is embracing AI and maintaining transparency about its usage. AutoAbode use automation as part of their product offering and to support circuit design. Viper Innovations have embraced AI in their research team, and are using it in engineering and policy making, and Ed Surman said he had been using machine learning knowledge management tools for several years. Others in the room expressed concerns about skill shortages and mentioned how AI and 3D laser scanning could potentially replace experts. Mary Lee suggested that many big law firms are intimidated by AI and how it could impact their business, whilst Jess Brewster suggested some in the arts are concerned about the automation of creative development and as the conversation turned to the value of replication in AI, especially when it comes to being a celebrity, Jess Brewster suggested that AI could further enhance the value of celebrities by replicating their style, for example, musicians who use their voice and style to create AI-generated songs. Kim Jones talked about the human impact of AI, and questioned how it should be used for ideation and Jess acknowledged the potential for human and AI collaboration. She mentioned PI AI, a personal, voice-enabled AI Chatbot that can adapt to the mood of the person engaging with it. As a charity, costs are a major concern, making the use of AI to automate marketing an attractive prospect.

Tom Henson delivered a presentation covering the key stages of the development of AI – machine learning, neural networks, deep learning - and offered a brief history from ENIAC, the first digital computer in 1945, to ChatGPT, launched in 2022, gaining 1M users in 5 days and 100M users in 2 months. Now Microsoft are pioneering the adoption of AI, introducing Copilot into their 365 suite of products, and investing in OpenAI, the architects of ChatGPT. Peter Quintana emphasised that ChatGPT should not be used in place of human activity, but rather treated as a colleague, or a co-pilot, that would then give humans more time to think and do the things that AI is not (yet) capable of. He offered a short demonstration of ChatGPT as a tool to support strategic thinking and talked about ‘prompt engineering’ – question design – the careful construction of questions that set the context for ChatGPT to mine its knowledge base and provide answers that support a business’ objective.

The group discussion began with the suggestion that there is an on-going ‘arms race’ for dominance in AI between the tech giants at Microsoft, Amazon, and Google, to which people pondered whether Microsoft had already won. Peter said that the head of OpenAI had requested a meeting with US Congress to ask them to regulate AI, and everyone agreed that legislative bodies and the courts needed to catch up, particularly when it came to copyright issues. The risks of AI replacing some functions, such as legal counsel, was brought up and it was suggested that the profession might need to adapt, perhaps by providing emotional support for people navigating AI-led court cases.  Julian Lee asked if anybody would accept legal advice from an AI bot as AI can already be used to summarise lengthy email chains of evidence and could eventually provide a legal opinion or summary to take to a legal expert. Mary Lee raised the very real question over liability: if AI is providing advice, where is the accountability, and what then is the impact on Professional Indemnity insurance? 

Concerns surrounding the security of data, following the introduction of Microsoft’s Copilot, and the dangers of sharing sensitive business data with ChatGPT were highlighted.  Tom pointed out that Microsoft Teams is now AI enabled, as is Dynamics CRM, and that these can be linked to maintaining a record of conversations over Teams, something that previously would have had to be entered manually into the CRM.

Customer service was identified as an area where staff turnover could be an issue and where AI could be deployed to empower and train new staff.  AI could create quicker response times and help manage difficult customers. There is also the potential of using AI to customise the user experience on a website, increasing chances of conversion. Kim Jones mentioned the challenge of maintaining the company's values in AI interactions, which could be achieved by training the AI to respond in a certain tone of voice. Peter then proposed a potential restructuring of customer service, where support is done by AI, and human agents are redeployed to other tasks, like visiting loyal customers. The impressive capabilities of AI-generated videos and how they could be tailored for individual customers was also discussed.  

The group pondered the necessary technology needed to fully enable AI, such as access to call logs and emails, which should be logged and acquired to feed into the AI. Miint Marketing are quickly able to create unique content for every single email, resulting in open rates as high as 95%. They acknowledged that without the correct foundations in place, AI will become difficult to embrace. 

Concerns were then shared about the impact of AI on education, and James Taylor was worried that AI could start making things too easy, which would result in less skilled graduates. Peter pointed out that primary schools are embracing new technology, but secondary schools were not. Everyone expressed that the school system needs to be changed, especially as exams are still largely question-based. This further created concerns about the implications of AI for industries requiring high levels of qualifications, like law or medicine. Mary said that if students could use AI to cheat their way through school and university, then it would pose significant issues for certain professions. To counter this, the group considered the potential of AI, VR, and student technology to provide personalised education plans tailored to the individual’s learning needs. Everyone agreed that the recruitment process would need to shift in response; instead of CVs, physical assessments and task-based evaluations could become more relevant.

To manage the challenge presented by the rapidly growing number of AI tools - new tools are being released almost daily - Rhoda suggested Future Tools (https://www.futuretools.io/) as a useful repository of AI tools for research. 

It was agreed that businesses that fail to adopt this technology risk being left behind, and staff may seek opportunities in organisations that are embracing AI or automation to eliminate the boring and repetitive tasks and improve their employee wellbeing. Overall, the group expressed excitement about the opportunities AI presents and it was generally recognised that AI will be key for improving productivity and efficiencies. However, everyone also acknowledged that there were risks. They appreciated the speed of change and its impact on society, from employee wellbeing to the transformation of qualifications and training and ultimately agreed that AI could add real value as a collaborative tool rather than simply a replacement for human roles.