Expert Views on Advancing AI for Good
Exclusive interviews with Rumman Chowdhury (Humane Intelligence), Eric Loeb (Salesforce) & Anna Koivuniemi (Google DeepMind)
The treasure trove of insights presented at the 2024 AI for Good Global Summit, which brought together global experts to explore AI’s role in advancing the Sustainable Development Goals (SDGs), includes a series of one-on-one interviews with leading AI authorities.
In the “Views from the Summit” series, panelists expanded upon the critical themes of their on-stage presentations, dissecting the challenges and solutions in developing responsible AI governance, ethical frameworks, innovative policymaking, democratizing AI, combating biases, promoting sustainability, and fostering AI literacy.
Don’t miss these three standout conversations:
- Rumman Chowdhury, CEO of Humane Intelligence and member of the US Science Envoy for AI, offers her perspective on the surge in AI governance initiatives and the importance of inclusive global discussions.
- Eric Loeb, Executive Vice President of Government Affairs at Salesforce, discusses the pressing need for innovative governance and sustainable practices in the rapidly evolving AI landscape.
- Anna Koivuniemi, Head of Google DeepMind Impact Accelerator, highlights AI’s transformative impact on global sustainability efforts, particularly in combating climate change and promoting equitable access to technology with startups like GraphCast and Alphafold.
Rumman Chowdhury, CEO of Humane Intelligence and member of the US Science Envoy for AI
Dr. Rumman Chowdhury has made a name for herself by promoting AI development while insisting on proceeding ethically. She emphasized the importance of this balanced perspective at the this year’s Summit.
In one of the main stage panels, Rumman Chowdhury described the Summit as two conferences: the glitzy center stage featuring high profile panelists and the behind the scenes workshops where practitioners were working to solve problems.
I feel like there have been two conferences happening over the last few days. There is one conference that is this room with the neon lights and the glitz and the glamour and the big screen and Nick Thompson from the Atlantic interviewing people about their opinions. There’s another conference happening. It’s mostly on the third floor, I believe… There are no neon lights. There’s no applause at the end of it. But they’re putting their heads down and they’re working together and they’re solving problems. I’m really heartened to see that happening.
Evolving Governance Alongside Technology
In her one-on-one interview, Dr. Chowdhury reiterated this point. She discussed the surge in AI governance initiatives, highlighting the involvement of multi-stakeholder groups that bring diverse expertise to global discussions. She emphasized the necessity of governance for fostering trust and ensuring the responsible use of AI technologies. Specific examples include Brazil’s comprehensive AI bill, the EU AI Act, and various US state-level laws.
Dr. Chowdhury also underscored the importance of including the global majority in governance conversations, noting the current focus on capacity building and data collection in these regions. She stresses that governance must evolve alongside technological advancements and argues for a balanced approach that fosters innovation while maintaining responsibility. Her optimism extends to the belief that practical implementation of governance frameworks is progressing positively, moving beyond high-level discussions to tangible actions.
One can innovate and frankly can only innovate if you do it responsibly… I think people are moving into practical matters of what it looks like to implement governance and moving past sort of the high-level narratives.
Eric Loeb, Executive Vice President, Government Affairs, Salesforce
In this interview, Eric Loeb provided a compelling case for the pressing need for innovative governance and sustainable practices in the rapidly evolving field of AI.
Governance Challenges and Innovative Solutions
Addressing the complexity of AI governance, Mr. Loeb emphasized the need for innovative and adaptive policymaking.
Mr. Loeb called for a balanced approach combining voluntary commitments and regulatory frameworks to ensure responsible AI development. He stressed the importance of transparency and trust, suggesting that AI should be directed intentionally, a concept Salesforce refers to as having a “human at the helm” rather than a “human in the loop.” This approach means that humans set the goals, define the strategies, and oversee the general trajectory and application of AI, but do not necessarily intervene in every specific decision.
Salesforce thinks of it as a ‘human at the helm’. A ‘human at the helm’ means you’re intentionally directing the trajectory of AI and giving it a strategic plan for how you want to be using it.
Immediate Action on Bias and Sustainability
Mr. Loeb identified bias and sustainability as critical issues requiring immediate attention. He urged stakeholders to focus on responsible design decisions to mitigate bias in AI systems.
On sustainability, Mr. Loeb pointed to the escalating energy consumption of AI systems as a significant concern. Salesforce is proactively addressing this through the development of AI sustainability policy principles and green codes. These initiatives aim to guide developers in making incremental yet impactful changes to enhance energy efficiency.
Additionally, Salesforce’s Net Zero Cloud, powered by AI, helps organizations identify and manage their emissions more effectively, aligning with both ESG reporting requirements and broader sustainability goals.
Navigating the AI Transformation
As AI continues to transform society at an unprecedented rate, Mr. Loeb’s insights underscore the imperative for collaborative, transparent, and sustainable governance. To ensure that AI’s benefits are equitably distributed and its risks effectively mitigated, we need to address immediate issues such as bias and sustainability.
Anna Koivuniemi, Head, Google DeepMind Impact Accelerator
In this conversation marked by enthusiasm and urgency, Anna Koivuniemi emphasized the pivotal contributions AI is making to global sustainability efforts, particularly in combating climate change and promoting equitable access to technology.
AI Driving Progress on SDGs
Koivuniemi highlighted a significant surge in AI applications tailored to SDGs, citing reports from the International Telecommunication Union (ITU) and Google that reveal a 300% increase in such use cases. She advocated for the democratization of AI tools and resources to ensure inclusive benefits.
Climate Change Mitigation
The conversation turned to climate change, where Koivuniemi presented three illustrative examples of AI-driven initiatives:
- GraphCast: An AI-based model providing accurate 10-day weather predictions. “AI in climate is a big opportunity,” Koivuniemi noted, referencing research suggesting AI’s potential to help mitigate 5–10% of global greenhouse gas (GHG) emissions by 2030 — the equivalent of the total annual emissions of the European Union.
- Optimal Power Flow (OPF): This initiative aids in the integration of renewable energy sources into the power grid. It optimizes power flow and adjusts grid predictions, supporting the shift towards electrification and renewable energy.
- AlphaFold: Initially developed to predict protein folding, AlphaFold is now employed in scientific research to develop enzymes that can break down plastics, highlighting AI’s role in driving biological advancements.
Tackling Inequality and Bias
Koivuniemi did not shy away from addressing the challenges posed by AI, particularly the risk of exacerbating inequalities through biased algorithms. She outlined DeepMind’s comprehensive internal practices to ensure responsible AI development and deployment, including rigorous evaluations and benchmarks. Additionally, Koivuniemi stressed the importance of fostering a diverse talent pool in AI development to better understand and mitigate biases.
Building AI Literacy
A strong advocate for AI literacy, Koivuniemi highlighted initiatives like Experience AI, a course designed for teachers of 11 to 14-year-old students, aimed at demystifying AI and educating young minds about its benefits and risks.
Conclusion
These interviews just scratch the surface of the expert views shared at the AI for Good Global Summit. Don’t let the details of current challenges and the examples of practical solutions that this interview series helps highlight get lost in the abundance of information presented at the Summit. Stay tuned for more of our highlights, insights, and in-depth analysis from the 2024 AI for Good Global Summit.
Learn More
Watch the AI for Good YouTube channel for Main Stage talks, highlights, and interviews from the AI for Good Global Summit.
Read the most recent edition of the UN Activities on AI Report, which showcases 408 AI cases and projects run by the UN system covering all 17 of the SDGs.
StartingUpGood supports fresh entrepreneurial approaches to social impact. FOLLOW US on social media:
Check out SDGCounting for the latest news on tracking the progress of the Sustainable Development Goals. #SDGs #GlobalGoals.
Disclaimer: This article uses YouTube transcripts and ChatGPT-4o to help summarize and proof-read. All content for the article was hand-curated and checked for quality.
