I have the tremendous pleasure to live, work and recreate on the unceded traditional territories of the xʷməθkʷəy̓əm (Musqueam), Sḵwx̱wú7mesh (Squamish), and səlilwətaɬ (Tsleil-Waututh) Nations. Like many of you, I replenish my mind, body and soul through the practice of running. Recently, I ran along the paths of Kitsilano and False Creek in Vancouver. While I don’t always “habit stack” today I was. On this run, the rhythmic sound of my footsteps was accompanied by the The Ezra Klein Show. This particular episode, marking the one-year anniversary of ChatGPT, delved into the profound implications and advancements in artificial intelligence. An intellectual exploration, albeit with a political-bent that can sometimes give this show a lefty-media reputation. All of these things aside, as I pondered the insights shared by the tech journalists Kevin Roose, Casey Newton and Ezra Klein himself I was struck by the change that we are all being forced into in our work lives and eventually into our personal lives.
Reflecting on the past year since ChatGPT's release, the the journalists highlighted how OpenAI has remained in the spotlight, despite recent corporate upheavals. Beyond the headlines though, the real story lies in AI's unrelenting progress and its increasing integration into our daily lives. This resonated with me as I admired the seamless blend of nature and urban development around me, a metaphor for the integration of AI in our lives.
Roose and Newton discussed the burgeoning market for AI social companions and the cautious approach of many companies. This cautiousness mirrors our own hesitancy to embrace new technologies. As I navigated the flat path among the picturesque beauty of the sea wall, just as the sea wall brings together diverse groups of people, AI’s ability to provide basic communication and engagement is something that can benefit all people.
The conversation then shifted to AI "interpretability" research, a field striving to make AI's decision-making processes more transparent. As I navigated the paths frequented by joggers, cyclists, and walkers, I considered the importance of balance in AI development. Just as the sea wall is maintained to cater to different users, the development of AI requires a balanced approach. It's about ensuring that as AI evolves, it does so with consideration for ethical implications and societal impact. Trust and accountability are crucial for society to benefit from AI systems. While these words have become diluted by the sheer repetitive nature of their mention in academia and the media, this is a critical area for research and policy. This is also an area where I think there is a need for social learning and education.
Here is one example of an area where trust and accountability can be implemented but it requires some involvement and leadership from corporations building the technology and government bodies looking to regulate the industry. The academics behind the blog “AI Snake Oil” argue that Reinforcement Learning with Human Feedback (RLHF), which is when humans work side-by-side with the machine learning algorithms to defend against bias and misalignment in how the models predict the information that is being given to the user.
As the role of government becomes crucial there is a key role for corporate leadership. The majority of the major AI advancements have come from the private sector. No question that these companies had support from Academic partners. However, this struck a chord with me, as I considered how public spaces like Vancouver's sea wall are maintained for the benefit of all. Similarly, AI development requires a balanced approach, ensuring public interests are safeguarded. What does that mean? It means that access to the technology, sharing the technology with academic and third-parties who can validate that safeguards are being met. Further, we need to start measuring AI’s impacts on labour productivity, energy intensity, wage impact etc… measure the impact of AI as we go through this transformative change is going to be crucial.
As my run concluded, so did the podcast, leaving me with much to ponder about AI's role in our society. The thoughtful and provocative discussion brought forward by the podcast, combined with my runner’s high, underscored the importance for me of understanding and engaging with AI. As we stand at the cusp of a new era in technology, it's imperative that we, as individuals and professionals, stay informed and proactive in shaping an AI-integrated future.
At Green Analytics we are focused on developing safe, value-driven and transparent AI. Further, we work with other AI technologies and in so doing try and learn as much as we can about these technologies and the companies behind them.
If you’re interested in a particular area of AI, as it relates to solving environmental problems. Reach out to us. We are always looking to meet new collaborators, partners or clients.
Green Analytics’ new venture, Sunterra Analytics, will help bring data and operational intelligence to the agricultural sector. The partnership between Sunterra Farms and GA will serve both companies in supporting new and existing customers and their agricultural operations, through leveraging data and custom monitoring solutions.
Green Analytics has achieved a major milestone in forestry innovation, using high-resolution LiDAR data to produce detailed outputs that accurately measure individual trees, canopy cover, and forest delineation across Canada. This breakthrough will enhance our work on natural asset inventories and provide high-value data that municipalities and industries can utilize for environmental planning and sustainability efforts.
