Demand for AI specialists is predicted to rise by 40 percent over 5 years: IBM's Sriram Raghavan

AI for weather and climate is one area that IBM is actively exploring. In collaboration with NASA, we are developing an AI foundation model that combines IBM’s AI leadership and NASA’s Earth science expertise to offer significant advantages over existing technology. This foundation model aims to have several advantages over current AI models: greater data diversity, faster inferencing time, greater accessibility, and improving the accuracy of forecasting for other climate applications. A few sample applications of the model could include inferring high-res information from low-res data, identifying conditions conducive to everything from airplane turbulence to wildfires, and predicting meteorological phenomena.

Are you concerned about the intersection of AI and Black Swan events? What are some of the safeguards you think might be useful?

While AI poses risks, just like any technology, it can also be critical to preventing these kinds of events. Enhanced risk management, predictive capabilities, and real-time reporting are all ways that we can build more trusted and resilient AI models. However, the most important thing in understanding and assessing risk is information, which underscores the need for open innovation and information sharing. To this effect, IBM and Meta, along with over 50 organizations from across the globe, recently launched the AI Alliance to help advocate and support an open approach to AI innovation, including the development of open benchmarks, safety tools, and safety evaluation methodologies.

Given the amount of compute needed, and the number of industry and government players interested to join in, is sustainable AI a pipe dream? Why/why not?

Sustainable AI isn’t a pipe dream, but a priority for many businesses today: a recent IBV study found that nearly half of CEOs cite sustainability as one of the highest priorities for their organization. As AI adoption grows, we must proceed thoughtfully and adopt a sustainability mindset. IBM is already finding new ways to reduce energy and carbon when using AI, such as deploying energy-aware algorithms, seeking out energy sources that are less carbon-intensive, and developing more energy-efficient hardware. Creating more sustainable operations isn't just good for society - it's a business imperative. Therefore, clients are prioritizing efforts that optimize their operations and cost, while reducing their waste and emissions.

In terms of developing AI talent, what do you think countries, companies, schools, individuals should be doing right now?

It’s imperative that organizations make developing AI talent a priority, and IBM is already working with public, private and university partners to support these efforts. Demand for AI specialists is predicted to increase by 40 percent over the next five years, adding about one million new job opportunities to the global economy. At the same time, it’s important to ensure underrepresented communities are benefiting from – not being further disadvantaged – by AI advances. Equitable access is a priority for us, and to help close the global AI skills gap, this fall IBM announced a new commitment to train two million learners in AI by 2026. This includes expanded collaborations with universities and NGOs around the world, as well as launching new generative AI coursework through IBM SkillsBuild.