The Economics of AI: Resilience & Continuous Value


Enterprise AI can bring an undoubtable amount of value when applied throughout an organisation. However, it’s fundamental that one considers the full economic impact of AI, including the costs, to create a truly sustainable Enterprise AI strategy. Getting to grips with the economics of AI is the first step towards taking your vision of Enterprise AI and turning it into long-term value.

The first principle of AI economics as established by Alexis Fournier, Director of AI Strategy at Dataiku, focussed on introducing reuse and capitalisation culture to reduce costs. In this post, we’ll introduce the second principle of resilient AI that produces continuous value, and the key points you need to consider to achieve that.

In this constantly evolving environment, from technologies, tools, and people, to regulation, delivering continuous increased value to your company might be a challenge. While you may have succeeded in scaling, the question of whether your initiative will last will start to plague your mind. Will the things that you’ve built in the past and the work you’ve put in today be useful tomorrow? Will your assets remain useful? This is all a question of AI resilience, the ability to both deliver and be sustainable over time.

AI resilience can be broken down into the following three parts:

1. AI Technology Debt

Will your AI environment survive technological transformation?

Technology has rapidly changed and developed within the last 20 years, so much so, that the tech that we use today didn’t exist even five years ago. So, what does that mean for the technology that you use today to create your AI initiatives? Will you retranslate all that code every time there is a new development? This consideration is key when forming your holistic AI strategy.

2. Loss of Control

Will you be able to keep your AI under control?

As you scale up your data efforts, your team will naturally expand and your landscape will become more complex. So, how are you ensuring that there is a unified movement amongst all parties? Being able to have easy access to the latest version of your project is crucial. Being able to easily identify who has done what on which datasets will empower teams to communicate and collaborate more effectively. Data and AI governance shouldn’t be an afterthought and it’s important to get these systems in place early in your Enterprise AI journey.

3. Rogue AI

As humans, our culture and perspective on the world naturally shifts and changes over time. Rules and legislations that were enforced 50 years ago don’t cut it anymore when stacked against current views. This should apply to your AI as well. What measures are you taking to avoid the reinforcement of old biases in data?

Factoring each of these three points into your AI strategy will enable your organisation to keep your AI resilient and deliver continuous value.

Join Alexis Fournier at Big Data Analytics for his seminar, AI Economics: Resilient AI & Continue Value Creation. Alexis will further expand on the second principle of AI economics, delving deeper and offering solutions towards hitting the mark on AI resilience.