07/26/2024 | News release | Distributed by Public on 07/26/2024 05:19
As the Global Head of AI Enablement, my daily conversations center around using artificial intelligence (AI) responsibly to deliver trusted outcomes for CGI and our clients. But I don't need to be an AI executive to experience its vast reach. My children use AI tools to help them gather sources for their university research assignments, and my mother used AI tools to plan her hotel layover in Paris.
The pace of AI's advancement has already overtaken previous disruptive technologies like the Internet and mobile devices. According to our 2024 CGI Voice of Our Clients research, AI is a top business, IT and innovation priority. Across industries, CGI's research reveals that 26% of executives interviewed are implementing traditional AI-up 10 percentage points year-over-year-and 79% are investigating or conducting proofs of concept for generative AI. This is no surprise given AI's potential to accelerate optimization and growth.
While the AI opportunities vs. risks debate continues, organizations that understand AI's potential also acknowledge the responsibility that comes with its use. This means focusing on ways to responsibly integrate AI into the multifaceted dimensions of our daily lives.
We see many examples where AI is helping solve some of the world's biggest challenges. Here are a few:
There is growing interest in using AI to advance diversity, equity and inclusion (DE&I)-a topic close to my heart. I am passionate about advancing DE&I progress both at CGI and in my community. I coach women within and outside CGI, including the Canadian high school international women's business team.
AI's potential in the DE&I realm includes:
1. Understanding AI and questioning how it works
To make a positive difference with AI, or for that matter, any technology, understanding how it works is the first step. There are three important parts to AI: the design of the algorithms, the data on which the technology is trained, and the decision-making processes included in the models.
For example, if we look at gender diversity:
We can start by understanding the design step. A study by UNESCO reveals that women occupy only 20% of technical roles in major machine learning companies, while only 12% of AI researchers globally are female. However, according to the World Economic Forum's Global Gender Gap Report 2024, the concentration of female talent in AI engineering has more than doubled since 2016, signaling a positive sign of improvement in gender parity. This is just one example. Overall, there needs to be more diverse representation throughout the AI development life cycle. With generative AI's potential to enable customized learning, AI can be a positive force in further improving this curve.
Underlying training data also often has an inherent bias. For example, training data for a recruitment algorithm that overrepresents males or specific demographics may hinder diverse hiring. AI can also be used to increase racial diversity in hiring. There is currently a lot of work underway to improve AI models to reduce racial biases. For example, Google Research Center is releasing the Monk Skin Tone (MST) scale for Computer Vision systems to understand human skin tones better and help the broader industry reduce biases.
Finally, human decision-making, or human-in-the-loop machine decision-making, must be based on trusted data or an awareness of underlying biases. My colleague Diane Gutiw shares how we closely monitored a chatbot in the healthcare space to ensure it yielded accurate and in-context results.
2. Having a positive human-centric vision
Human-centric design is another area where AI offers immense potential. It is about enabling the intrinsic ability of humans, not taking away from it. For example, AI speech translations, text-to-voice and vice versa can improve accessibility and break down communication barriers tremendously. Generative AI tools like ChatGPT, Perplexity AI, or Microsoft's Copilot for Microsoft365 are already improving productivity and creativity, making work more motivating, fun and enjoyable for many individuals.
Developing a human-centric positive vision is essential to ensuring positive outcomes. My colleague Fred Miskawi outlines what this could look like in his blog on human + artificial intelligence.
3. Building an ecosystem of trusted partners
To realize the potential of AI, working alone is not an option. Organizations need trusted partners, lawmakers and governments to operate in sync. Think about partners that deeply understand the technology and the evolving regulatory scene and ones that have a vast, trusted ecosystem of partners of their own. For example, local and international consortiums of government officials, regulators, leaders and scientists from research institutes and large technology firms can facilitate the ethical and responsible use of AI.
Where do we go from here?
We have real opportunities that can unite us toward a common goal of making AI a positive force for people's lives and work. AI is here to stay, and it is up to us to intentionally envision and shape a positive future. Let us start by understanding AI, asking the right questions, having a human-centric vision for using the technology, and collaborating with an ecosystem of players.
Contact me if you'd like to discuss more.
Christina leads CGI's global AI Enablement Center of Excellence where she focuses on AI strategy and offerings and drives the company's mandate to enable the responsible use of AI.