AI v the CTO

In this excerpt from my conversations with 100 CTO’s I cover the topic of AI. It’s a fascinating topic and one which is having an impact on the role of the CTO currently. The CTO’s I spoke to had mixed opinions around the future of work with AI. In the main they were optimistic and of course curious and interested in how AI can benefit their team and impact the business. However, there were some words of caution around key concerns.

The CTO’s I spoke to are increasingly having to look at integrating the technology into existing systems as well as understanding how it can enhance customer facing products, or indeed help create new ones. It is a fine line for them to walk. On the one hand they need to get to grips with the limitations and risks: managing C-suite expectations, and on the other hand having to understand the enhancements it can bring, identifying areas where AI can drive innovation, efficiency, and competitive advantage. AI relies heavily on data, making data strategy a critical component of the CTO's role. This includes data governance, ensuring data quality, and establishing robust data pipelines. Ethics of course plays a huge consideration and the rise of AI necessitates a new set of skills within the technology team. CTOs are finding themselves increasingly responsible for collaborating with HR to create training programmes and career development paths that align with the organisation's AI strategy.

There’s a general feeling that for better or worse it is here to stay, it is in our faces loud and more persistent than previous iterations (driven mainly by the very public 2023 launches of generative AI). I asked them for some of their strategic insights around the rise of AI in general and this is a snap shot of what they said:

• The winners will be those who can get past the hype-train and harness and apply it in a compelling manner.
• It’s worth remembering that applying AI in your business doesn’t mean it has to be a customer facing feature.
• There is a lot that Gen AI can do to help your business be more effective. Coding assistant tools like GitHub Copilot or Amazon’s data analytics agent Quicksight Q are helping people do their jobs better every day.
• Many sales and marketing departments are already taking advantage of content generation and agent tools in their daily routines.
• At this point, companies can gain an edge simply by embracing these new tools.

Of course, one of the challenges CTO’s face is managing the expectations of the C-Suite. When I asked them for their advice on how to influence in this area, these are some of the points raised.

What advice do you have for managing the expectation of the board when it comes to AI adoption?

• We’re at the peak of heightened expectations. For example, the Head of Marketing or the CEO who gets a full sales pitch for Gen AI then passes that heightened expectation heavily onto the heads of those in charge of tech transformations. However, the foundations need to be in place first, and all too often they are not. ?
• Get your data together. Make sure the quality is there, make sure it is being cleaned, and there are testing models in place.
• Understand that implementing this technology requires quite a big investment so be careful to avoid the unicorn that is a donkey wearing a costume.
• If you are going down this route, early on, pull together a paper for the board. Describe the potential benefits and draw backs. Introduce healthy scepticism early on to counter the sales pitches and ensuing expectations.
• Use the interest around AI to drive a resurgence on the importance of data management and governance.
• Manage expectations and educate to mitigate AI-washing – it is not the solution to everything!
• First mover advantage isn’t always the best. Customers are getting jaded and more cynical by the day.
• Cost savings from using AI agents might be great but to what end? There is a growing backlash from customers around poorly implemented AI in many areas.
• It will take some discipline to find product fit. You should have a deep understanding of your business and not just the technology aspect of it. Enough to know when not to throw AI at everything.
• Sometimes, good old software code is cheaper, faster, and better understood.

Collaboration across business functions and teams should increase when introducing new technology. CTO’s are managing the transition from traditional IT infrastructure to AI-enabled systems, often involving cloud computing, big data analytics, and IoT. Because implementing AI often requires significant changes to business processes and workflows, resistance to change from employees can hinder the successful adoption of AI technologies. CTOs need to lead change management efforts, including training and communication strategies, to ensure smooth transitions and buy-in from all levels of the organisation.

When I asked them what they saw as the main considerations for AI transformation, they raised a few different points but cost was right up there:

• Tailoring AI models to your specific use case can get very expensive very quickly. Both salary as well as infrastructure costs continue to be high as we’re all competing for the same finite talent and resource pool.
• Carefully consider if your business should be creating specialised AI models and systems from scratch. Can you build on top of what most major IaaS players’ offerings instead, as is more often the case? Take the time to survey the landscape and understand what your options are.

Another area that was raised was continual learning:

• No matter what type of CTO you are – either more hands on or more leadership/strategy centred – staying up to date using a combination of google news queries and reddit is important. Definitely look beyond Forbes articles and keep having the conversation about AI.
• Continue reading up and network as I am sure we’re all doing already. But also, don’t forget that there is a thriving research network in academia and applied sciences that you can tap into to get ahead in the game. For example, in London, there’s the RAAIS that brings people from both sides (businesses and researchers) of AI together each year at their annual summit and talk about all things AI and their applications. So, find out where your nearest networking hub is and go from there!
• I would also encourage CTOs to keep an eye on legislation changes in the sector and geographical regions they operate in. Law making has always been slower than technology and it’s no different for AI tech. That said, being ill informed of legal requirements can have a significant negative impact.

One area we haven’t yet discussed, but it’s very much on the minds of the CTO’s I spoke to was how AI systems can inadvertently perpetuate or even exacerbate biases present in the training data.

The impact of growing a diverse team can’t be stressed enough here and it is an area many data teams struggle with. A portion of CTO’s raised the point that some algorithms can be life changing and if bias creeps in that’s obviously going to be detrimental to some sections of society. Other CTO’s raised the point that more diverse teams make better decisions and that there is a real need to ensure that any data used to augment LLM’s needs to be screened for bias. It is the responsibility of the CTO to ensure that any algorithms are regularly audited for bias. In all cases there was a real view that it was an opportunity for business strategy and data strategy to be brought into alignment if it wasn’t already. Essentially CTOs must address ethical considerations and ensure that AI systems are designed and implemented responsibly. This includes tackling issues such as bias in AI algorithms, data privacy, and the societal impact of AI technologies. Establishing frameworks for ethical AI use and compliance with regulations is becoming a key part of the CTO's mandate.

A Final Word on Risk Management and Security

AI introduces new risks, such as cybersecurity threats and algorithmic vulnerabilities. CTOs are finding themselves developing strategies to mitigate these risks and ensure the security of AI systems. They need to find ways to implement robust cybersecurity measures and continuously monitor AI systems for potential threats.

In summary, AI is significantly reshaping the role of the CTO by emphasising strategic integration, data management, ethical considerations, talent development, innovation, cross-functional leadership, operational efficiency, and risk management. CTOs are having to adapt to these changes to effectively leverage AI technologies and drive their organisations forward in the digital age.

?