In a recent earnings call, Brian Chesky, CEO of Airbnb, revealed that rather than just providing a service or product, the brand envisions itself as a global travel community and the ultimate host. By building on top of base models such as GPT-4, Airbnb aims to create a robust AI ‘concierge’ which understands users on a deeper level. The tuning of the AI model is based on customer data, allowing for personalized recommendations and matching that goes far beyond traditional search queries.

The company is investing in initiatives such as the so-called Host Passport to learn more about guests and hosts; thereby enabling it to provide tailored accommodations and travel services. This personalized approach extends to customer service, too, with AI augmenting agents and improving consistency, speed, and cost-effectiveness. By integrating AI into its product, including the immersive and multimodal capabilities of GPT-4, Airbnb plans to enhance user interfaces with rich media experiences.

But Airbnb is far from alone in applying a sophisticated lens to new technologies. This new frontier of large language models (LLM) and AI agents provides huge opportunities, but brands must take care to promote inclusivity, equity and diversity within this ecosystem. They must ensure that women have an active seat at the table and are helping to shape the future of AI  – despite the fact that the tech world remains dominated by men.

History has proved that giving women active roles drives positive outcomes. Role models such as Fei-Fei Li, a Chinese-American computer scientist and co-director of the Stanford Institute for Human-Centered Artificial Intelligence, is considered one of the most influential women in AI. Li and a few others shine a spotlight on the valuable contributions that women can and are making in the field. However, the underrepresentation of women in the industry remains a pressing issue.

Statistics show that only 22% of data and AI professionals in the UK are women, and a mere 8% contribute to leading machine learning conferences. This lack of diversity not only hinders economic equality but also perpetuates bias within systems.

This is not only an issue of economic equality, but also about how the world is designed and for whom. Mounting evidence suggests that the under-representation of women and marginalised groups in AI results in a feedback loop whereby bias gets built into and amplified by machine learning systems. Addressing the gender job gap in AI is the first step to ensuring that our technology works for all of society.

Certainly the application of the latest technologies requires a thoughtful and intentional approach.  One important consideration is the up-front integration of diverse, inclusive and unbiased data sets during the creation of AI agents and LLMs. If these tools do not represent all individuals, such as women and minorities, equality gaps can quickly widen. Employing data collection from a wide range of sources, plus augmenting data when needed, can help to boost overall diversity. 

One critical way to accomplish this goal is to ensure that there are intrinsically diverse teams working on AI initiatives. This means implementing hiring practices that prioritize inclusion - and creating KPIs that ensure these practices are measured and met. Companies should consider placing women and individuals from minority groups in leadership positions, overseeing and developing AI systems that bring in their unique insights, experiences and perspectives. This approach puts these important voices on-the-ground - identifying and addressing biases that might be inadvertently included in AI systems.

We can’t ignore rising ethical considerations as AI becomes more and more prominent. Google’s CEO, Sundar Pichai recently addressed this very topic in the FT, describing AI as ‘the most profound technology humanity is working on today’, and writing: “Given these high stakes, the more people there are working to advance the science of AI, the better in terms of expanding opportunities for communities everywhere”, stressing the importance of building such technologies responsibly and making sure at ‘as a society we get it right’.

We talk about ‘winners’ and ‘losers’ within the AI race, establishing and following ethical guidelines is the responsibility of everyone playing in this sandbox in order to ensure not only that such tools and systems are safe, secure and private, but that the landscape itself is trustworthy. Creating collaborative cultures, where sharing best practices is part of the development process, is also part of ensuring a responsible, trustworthy technological environment.

Ensuring diverse data inputs, hiring diverse teams, prioritizing ethical practices and actively collaborating with others in the field are just a few practices that help AI and LLMs grow to become instruments of positive change for brands. Bias in these powerful tools begets more bias, so making inclusivity, diversity and equality part of their foundation is essential to ensuring AI technology serves the needs of all people in our society; not just a select few. The brands which lead in this area can make AI a positive agent of change and a force for good within society more broadly.