- Google DeepMind's Lila Ibrahim urges AI regulation to balance safety with innovation benefits
- She calls for global collaboration among governments, companies, and stakeholders in AI oversight
- DeepMind partners with Hyderabad's crop institute to enhance sorghum research using AI tools
Google DeepMind Chief AI Readiness Officer Lila Ibrahim said artificial intelligence regulation must minimise risks without limiting the technology's potential benefits, calling for collaboration between governments, AI companies and other stakeholders to ensure its responsible development.
In an interview with NDTV, Ibrahim said AI development required safety and responsibility to be built into the process from the outset, rather than treated as an afterthought. "We have to figure out how do we regulate AI, but how do we do so in a way that's actually right-sized for the technology," she said.
Ibrahim said regulating AI could not be the responsibility of any single company or country, given that models were available to people worldwide. She advocated an application-specific approach to regulation, focusing on how AI was used rather than treating it as a standalone technology.
Ibrahim highlighted Google DeepMind's partnership with the International Crops Research Institute for the Semi-Arid Tropics, headquartered in Hyderabad, to advance research into sorghum, known as jowar in India.
Sorghum provides food, fuel and livestock feed for more than 500 million people, she said. However, the crop has received comparatively limited research attention, with much of the existing work focused on varieties grown in Western countries rather than those indigenous to India and Africa.
ALSO READ: 'No Degree Required': Google DeepMind's Rs 11 Crore Job Listing Sparks Tech Debate
The partnership aims to develop an open-source set of AI tools to support research into sorghum, including the development of disease-resistant and drought-resistant crops.
Ibrahim said the project was still in the research stage, with crop breeders being consulted regularly to ensure their requirements and challenges inform the work from the beginning. The initiative would involve developing a catalogue for crop breeders, followed by further work to develop seeds, plant them and learn from the results.
She said AI could help accelerate research that might otherwise take more than a decade, while stressing that collaboration would be critical to ensuring the benefits reach farmers.
On AI adoption in education, Ibrahim said the technology could enable personalised learning, provided it complemented teachers rather than replaced the learning process. She cited Gemini's Guided Learning feature, which helps students work through problems step by step instead of simply providing answers.
According to Ibrahim, an eight-week trial in Sierra Leone found that middle-school students using Guided Learning with teacher support advanced by the equivalent of a year and a half in mathematics.
She also cited a pilot programme in Northern Ireland in which teachers reported saving 10 hozurs per week through the use of Gemini. However, Ibrahim cautioned against evaluating education solely on test results, stressing that the focus should be on whether students were actually learning. “I would love it if we really focused on our kids learning, not are they just passing a test,” she said.
She added that teachers had an important role in helping students use AI as a learning tool rather than as a shortcut for completing assignments. When it came to AI's relationship with employment, Ibrahim said people entering the workforce needed to prepare for jobs and responsibilities that would evolve as technology advanced.
Drawing on her own career as an electrical engineer who worked on the Pentium processor before moving into artificial intelligence, she said workers needed to keep developing their skills and learn to use AI as a tool.
She identified contextual understanding, curiosity and problem-solving as important capabilities, adding that organisations should identify the distinctly human contributions employees bring to their work.
AI could assist with background research and routine tasks, she said, but people would still need to bring judgement and an understanding of the people and organisations they worked with.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.