It’s now been nearly two years since ChatGPT first came into the public consciousness, amid a stream of headlines all over the world. While many were excited about the potential of putting artificial intelligence within easy reach of the public, the reaction wasn’t universally positive.
Across a number of industries, many employees have concerns that AI could do their jobs much faster, efficiently and cost-effectively than they could do themselves, putting their employment or even their entire careers at risk. According to SurveyMonkey, 60% of employees who use AI at work are concerned about the impact on their jobs, compared to 35% of those who don’t currently use AI.
Some of these fears are, at least to a certain extent, well-founded. Goldman Sachs has reported that while generative AI could boost global GDP by 7%, it could expose as many as 300 million full-time jobs to being replaced by automation. But is AI going to be the dominant force in the business world that this suggests, or is there a way forward in which humans and AI can come together for the greater good?
How is AI being used in the workplace?
To understand how AI and employees might collaborate and converge in the future, it’s important to understand where AI has been commonly deployed so far. Four common use cases have emerged across most industries:
1. Workflow management and communication
AI is proving extremely useful at analyzing vast quantities of data throughout workflows, from processing times to order volumes, so that the workloads placed on employees can be optimized. The right tasks can be matched to employees with the right skills, improving the quality and quantity of work done, and cutting out bottlenecks in productivity. The same principle can be applied to peaks and troughs in demand, so that staffing levels can be right-sized to business needs at different times. That includes also communication between teams and employees and help with tasks such as translations, reviewing texts, creating summaries, etc.
2. Stock control
AI can proactively work to smooth out gaps in the supply chain and inventory for retailers, ensuring that any gaps in availability are plugged before they even appear. This is achieved through demand forecasting, alongside real-time monitoring of sales and supply-chain logistics, and is playing a vital role in satisfying customers and preventing them from shopping elsewhere.
3. Customer service
Chatbots powered by conversational AI are expanding in functionality all the time, and are gradually taking on more types of queries and issues from human customer service teams. This does not have to be considered a replacement for staff: instead, chatbots are taking on the more common and repetitive queries that can be answered simply and quickly. This relieves the time pressure on the customer service team, and frees up their skills to focus on problems that require longer and more complex resolutions. The travel industry has benefitted from chatbots in particular, as they have been able to deliver service 24/7 and in multiple languages.
4. Training
Many businesses are now using AI to identify skill or knowledge gaps that exist for individual employees, in order to pinpoint training opportunities they might need. In the world of finance, this concept is also being deployed at an organizational level, spotting areas within teams or departments where more expertise is needed, and addressing them accordingly with learning and development programs.
Are there benefits of using AI for workers and businesses?
When AI is deployed correctly and in the right places, it can be a real help for employees, rather than a hindrance or a rival. This delivers benefits not only for the workforce, but for the business as a whole:
Enhanced security and data protection
One of the areas where AI trumps humans is in being able to do things accurately, time after time. This can make a real difference when it comes to security and data protection, as it means that proactive monitoring and analysis of real-time data at scale can be conducted 24/7.
Increased productivity
Similar to the chatbots freeing up customer service staff, the ability for AI and automation to take care of the mundane and repetitive work means that any type of employee can apply manual skills where they’re needed most. This could be on a manufacturing production line or in frontline healthcare, but the principle is the same: AI can do what it does best, and humans can do what they do best, for maximum productivity and output all-round.
Resource management
The ability for AI to use large volumes of historical data and make more accurate and reliable predictions can be immensely valuable to any business. It allows them to forecast demand, whether it’s for the launch of a new product or extra footfall in the run-up to Christmas, and plan their resourcing accordingly. This means they can avoid the wasted cost of bringing in too many people, materials, products or machinery, or the deterioration of service caused by not having enough.
Faster data and insights
Not only can AI generate valuable business insights at a much greater level of detail, but it can do so much faster, as well. AI can be relied upon to trawl through huge volumes of data, and digest the information through its algorithms and machine learning capabilities. This can enable businesses to quickly spot new trends and patterns which they can then act on immediately, helping them maximize their agility, embrace new opportunities and seize first-mover advantage.
Routine task management
AI can help individual employees get more done in their days, just as much as it can support overall business productivity. Virtual assistants that are powered by AI are being used to automate and streamline many of the routine tasks that employees have to deal with every day, such as meeting scheduling, email management, and even finding important information or documents. The content creation capabilities of AI is also helping them draft and refine emails at a much greater speed than they could previously manage on their own.
What are the risks of using AI in the workplace?
Even if a business wants to support its human workforce, that doesn’t mean that AI should be shunned completely. The simple fact is that the scale of the change it can generate means it’s impossible to ignore it completely, and not adopt it in the areas where it can deliver the biggest gains. But just as organizations shouldn’t put all their eggs in the human basket, they shouldn’t put them all in the AI basket, either. Failing to combine the two properly exposes firms to a number of risks, including:
1. Being left behind by competitors
So many businesses are actively looking for practical ways to deploy AI. This means that any organization that isn’t using it to its full potential can be sure that at least some of its competitors will be. At a time when competition in marketplaces has arguably never been higher, AI can make a real difference to growth, revenue and market share. And at the same time, the employee efficiencies that AI delivers is increasingly playing a major part in staff satisfaction, and by extension talent retention and acquisition.
2. Failing to maximize efficiency
Not enough use of AI means that humans are still going to be overloaded with administrative work that they can’t keep up with, reducing the overall effectiveness and productivity of an organization. But too much use of AI, on the other hand, means that employees may be left twiddling their thumbs as the skills and experience they’ve built up over many years goes to waste.
3. Over-reliance on technology
Putting too much faith in certain AI solutions means businesses will suffer if those solutions don’t deliver on their potential. For example, in 2023, a number of researchers suggested that ChatGPT was ‘getting dumber’ and was delivering poorer results than it was previously. While the findings have been disputed by its creator OpenAI, it does demonstrate the fact that AI cannot be relied upon 100%: it is still at risk of generating irrelevant or biased information, and its ‘hallucinations’ can easily lead to dissatisfied customers, lost revenue and even legal issues.
4. GenAI tools and biases
GenAI can get as biased as humans can get too. So this will be an emergent challenge in the future - maintaining the quality of training data, as more and more will be generated by AI itself and some might even be artificially fabricated by humans with bad intentions. It is kind of a "technical sabotage". While masses are being "programmed" by using tools such as "propaganda", AI models can be also compromised by the nature of their environment and the data sources they have access to.
Use case: People working together with AI to create software
ON ONE HAND:
- Developers are increasingly using AI for refining code or generating boilerplate code.
- While AI does not excel yet at being enterprise architects or business analysts, it performs extremely well in identifying the right algorithms and detecting logical flaws, writing boilerplate code etc.
- Developers can focus more on being creative and productive rather than just coding. This shift unlocks humanity's potential for progress in engineering, as more intellectual capacity is now focused on science and engineering, thanks to AI assistants handling time-consuming tasks.
ON THE OTHER HAND:
- Since machines are now capable of understanding natural language, developers leverage this to combine the robustness of traditional programming with the ease of expressing ideas in natural language.
- While an engineering mindset remains essential, developers now have new tools to create smarter software faster.
- This means that we can produce RAG systems.
Use case: LLMs and research
One of the major transformations in human progress due to AI is its impact on research. Scientists can now digest and review more content with greater accuracy. They have assistants to see the unseen. Some game-changers in this field are:
- AlphaFold: Performs predictions of protein structure.
- BioMistral: An LLM trained on medical research papers.
- Motion2Vec: For surgical robots.
- GPT-4o: Multiple applications, including Medical Imaging, enhancing the accuracy and efficiency of medical diagnoses, drug developments
- MedPalm: Utilizes big data for medical diagnosis.
In summary: how can AI and humans work together?
AI is already having a huge impact on how people work. From enhancing communication, where people can now easily obtain meeting transcripts, summaries, or conclusions, and analyze large amounts of content to extract main ideas to removing language barriers, where people can focus more on what matters—ideas and making things happen—rather than struggling with language bias and cultural barriers.
As with so many things in the business world, getting AI right is ultimately a question of balance. Organizations that simply look at AI as a means of replacing human endeavor will not only be overlooking some of AI’s potential, but will also be letting valuable manual skills go to waste unnecessarily. The best way forward is to use AI and human intelligence together, devoting the right skills in the right areas, and supporting better outcomes for all.
If you feel you need help in striking this balance, then Ciklum can help as an expert in AI, machine learning and Process Automation technology. Find out more about our embedded AI and ML products, and how we can bring AI and your workforce together, here.
