DigitalTransformationTalk: Will the rollout of AI tools leave your organisation short of skills?

On 2 July 2024, Digital Transformation host Kevin Crane was joined by Ronald Gerlach, Chief Marketing Officer, Gaubert Oil Company; Martin Miller, Ex-Director of AI/ML Production Operations, Levi Strauss & Co; and Ayman Husain, Customer Engineering Leader, Google.

Views on news 

Google has announced plans to roll out AI “side panels” in its Gmail service, as well as in Docs, Sheets, Slides, and Google Drive that will showcase Gemini. Those panels will assist with your writing, summarize information, and suggest improvements or responses. The rollout has already started for select users and will begin for most on July 8. At the same time, browser maker Mozilla has announced plans to let users incorporate AI chatbots into the Firefox browser’s sidebar. Rather than allying itself with a single AI company, though, Mozilla plans to let users choose among offerings from ChatGPT, Google Gemini, Hugging Chat (an open-source alternative to ChatGPT), or Le Chat Mistral (from a French AI startup) to summarize packets of information.

These new tools will no doubt increase productivity and are unlikely to widen the AI skills gap. 

How to identify and bridge skill gaps

Users will pick up how to use these side panels just like how they did with browsers or emailing. The more so, as users can give prompts to these tool in their own language. On the technical side of prompt engineering there is going to be a widening skill gap but not on the users’ end. There are some caveats to GenAI’s use for summarising in an office context. Users need to see whether the tool sums up content through the proper lens for a particular domain. The new tools being rolled out may not be too much of enhancement for these use cases, but they are expected to be more commoditised and privatised. Also, a fall in productivity following implementation may be a telltale sign of an AI skills gap.

Although there are established processes, they need to be adapted to the circumstances of the company. AI and automation have been increasingly implemented in Learning Management Systems, which can enhance the effectiveness of training and be leveraged to support learning on the job. There are plenty of free courseware opportunities out there too – LinkedIn Learning has that too and there are YouTube tutorials. When looking for training technologies, it’s a challenge to pick one that is going to remain on the market in the long term rather than one that’s a dead and street in terms of technological development. Although many users can self-teach themselves bits, formalised, curriculum-based learning and training is the real answer. That said, what you learn at school will always lag slightly behind the technology that you must deal with in the real world, which calls for closer collaboration between industry and education. 

You can start out with a skills audit to learn what skills your workforce already has and compare to what you need for a new AI tool. Consultations with employees to learn what skills they are lacking are key too. Mentorship programmes will become common at workplaces, where the AI savvy can share their knowledge with the rest of the workforce. Good prompt engineering skills take domain knowledge and a level of understanding AI. Programmers, who have been using a programming language that is likely to become obsolete soon can leverage AI that prepares them for switching to a new one. Cross departmental knowledge sharing is key to bridging the AI skills gap.

To enhance this, every department must have an AI champion, who can get together regularly and become catalysts of change and information sharing. AI should get represented on the C level too. Set up regular meetings where information can be shared and document best practices. Establishing a centre of excellence is very helpful too, but you always need incentives to make sure that these best practices proliferate across the organisation, for example,w linking new procedures to better outcomes and an improved bottom line. 

The panel’s advice

• Young service engineers today often use their smartphones to identify parts and find out here to get them or watch TikTok videos to find out about procedures rather than asking their superiors. And they’ve picked these skills up without participating in formal training.
• Education on how to use AI must be rolled out everywhere, not just in tech hubs to create an equal opportunity for everyone to learn how to use the technology.
• Without data share across departments, AI share will be useless. Make data sharing look important.