In keeping with a new BCS Fellows Technical Advisory Group paper on the augmented intelligence economy, Brian Runciman MBCS, ChatGPT and BCS members collaborate to report on the prevailing view of members on AI’s current and near-future impact.
In the BCS ‘Digital in Business Life 2024’ survey, we asked some AI-specific questions in addition to the normal ‘planning-for-the-next-year’ style queries.
AI has been discussed in BCS journals and magazines for around 50 years – and in the last two or three it has suddenly risen in the general consciousness. What hasn’t come over quite so well is what AI is actually good for, what it can do, what it can’t do, what it can grow into and what it can’t.
The overall sentiment from this research is that AI holds significant potential across various sectors, particularly in automating routine tasks, enhancing decision-making, and improving the personalisation of services. Indeed, AI is already doing that. On the concerns side are worries about job displacement, ethics, and the need for robust governance frameworks.
The creator of the ELIZA chatbot, Joseph Weizenbaum, commented as far back as the 1960s that the subtlety of machines meant that even experts could be dazzled by their operation. Indeed, an issue today is the mythologising of AI’s capabilities. Weizenbaum also said, however, that once the “inner workings are explained in language sufficiently plain to induce understanding, its magic crumbles away.” That is a good goal for today.
AI impact 2024
We asked responders to comment on which sectors AI will have the greatest impact on in 2024. Then we asked ChatGPT to summarise the answers. We got a reasonable list, but perhaps without the flavour that real comments contain. Let’s look at both:
- Healthcare and social care: AI is expected to significantly improve personalised care plans, diagnostic processes, and treatment paths due to its ability to process and analyse large volumes of clinical data. It could alleviate pressure on healthcare systems and enhance patient experiences.
- Education: There is potential for AI to customize learning experiences, which could lead to better engagement and knowledge retention among students. However, it also raises concerns about academic integrity and the need for policy adjustments in educational institutions.
- Customer service: AI is seen as a tool to streamline service workflows, handle first-line customer inquiries, and improve the overall customer experience. There is a belief that as AI becomes more advanced, it could take on a larger role in customer service positions.
- IT and cybersecurity: AI is anticipated to be integral in transforming IT services, improving efficiency, and responding to cybersecurity threats. The sector may lead in demonstrating how AI can be adopted across other industries.
- Finance: There is a view that AI will enhance financial analysis and fraud detection. AI's ability to process large datasets could lead to more accurate market predictions and risk assessments.
- Content creation and media: Respondents expect AI to contribute to generative content, impacting journalism, social media, and digital content creation. This could lead to faster production of content but also raises issues about the devaluation of human creativity.
- Defence: AI's impact on the defence sector could be profound, potentially accelerating decision-making processes and operational efficiency, but it also poses safety and ethical concerns.
- Manufacturing and logistics: AI could improve efficiency and cost-effectiveness in these sectors by optimising production processes and supply chains.
- Software engineering: AI is helping to automate tasks in software engineering, increasing efficiency and potentially reducing the workforce.
- Legal sector: AI might be used to create legal documentation, which could impact the volume of manual legal work.
- General business: AI is seen as a tool that can automate many processes, support decision-making, and potentially cut costs across various industries.
In conclusion, as ChatGPT rather prosaically informs us: concerns are raised about the potential misuse of AI, the need for ethical deployment, and the importance of ensuring the technology is safe and unbiased. There's also an emphasis on the need for AI to adapt to and enhance human actions, rather than simply replacing them.
So what are some of the comments this was drawn from? The following are lightly edited verbatim comments…
Health
‘Pilots are already under way to assess the patient safety and accuracy in care settings, lack of health resources with increased demand drives the search for alternatives.’
‘Trending large amounts of data to look for patterns and correlation in populations that can discover unknown causes of illness.’
‘Using AI to simplify creating well designed services based on typical standards makes it so much easier for lower budget/public sector to create good services.’
Society, security
‘Current "AI" is more of an beefed-up expert system, rather than true intelligence, and politicians/civil servants don't understand it. They are highly likely to make rash, ill-informed decisions about AI, related technologies and legislation.’
‘With the continuing polarisation around the globe, associated wars, and the capability for bad actors attacking through cybersecurity, I feel that bad actors will increasingly use AI capabilities to enhance their actions, whilst governments and companies will need to use similar capabilities to defend against such attacks.’
‘The purpose of intelligence is to solve problems. No better unselfish priority than the things that make quality of life higher for those in need. Please don't just discover gunpowder all over again and set it to war and personal gain.’
‘Companies want to remove the risks of human inconsistencies in handling customer services and "streamline" complaints procedures. A server room also takes up less space than a call centre!’
‘Because AI is a marketing ploy and it depends on how many IT “professionals” are seduced by the spiel as to how much IT/IA resources is diverted into what is, as yet, a red herring!’
Education
‘My thoughts: AI will likely replace many traditional teaching roles and add more capabilities to remaining teaching roles. Other sectors, especially IT, health and defence, will also be impacted, but their peak impact might come maybe slightly later, e.g. 2025 and later. But for education, this will come rapidly already in 2024.’
‘Enable students to self-learn outside the classroom.’
For you
Be part of something bigger, join BCS, The Chartered Institute for IT.
‘AI will be able to personalise education according to each individual's needs. This will improve the learning experience significantly which could lead to more people wanting to improve their skills and creating more value to the workforce.’
‘AI already filters through all the dross of Google searches when looking for answers. Imagine applying that to a syllabus.’
‘There are a lot of positives for education providers, such as planning support and development of resources. However, it makes academic misconduct a greater risk but more difficult to identify.’
‘Negative impact with students becoming too reliant on it, believing responses are correct, and not practicing/applying their own capabilities. Outcomes will be skewed.’
Media and publishing
‘Creatives are always undervalued so I see this as easy pickings for people thinking they can automate away a "cost".’
‘Deepfakes and artificially created content are a massive influence on the general public who are not educated to see the manipulation.’
‘Popular AI at the moment is LLM, well suited to writing copy.’
‘Generating derivative material from existing material already published/broadcast. It'll make it easy to generate "more of the same".’
Customer services
‘The current hype on AI is about large language models. It feels that customer services will be the first to show a great impact. Customer services have a large base of natural language incidents, and a machine doesn't get offended with a frustrated or violent customer. If a customer support representative/engineer doesn't know a language, the Chatbot may be able to communicate perfectly with the customer in their native tongue.’
Finance
‘Repeatability of many financial regulatory and compliance tasks lends itself to an AI element.’
Skills
‘More helpdesks are using AI as the first point of contact with customers. Using AI for this purpose has enabled skilled staff to become the deep support, only required when AI is unable to resolve the issue.’
‘Copilot can help create code.’
