To bring to the forefront the significant contributions of women in the AI field, TechCrunch is embarking on a series of in-depth interviews. These interviews aim to shine a light on the remarkable women who have played a crucial role in the AI revolution. One such woman is Tamar Eilam, who has dedicated 24 years to IBM and is currently an IBM fellow, focusing on sustainable computing.

Unveiling the Women Shaping the AI Future

IBM Fellow Tamar Eilam: A Pioneer in Sustainable Computing

Tamar Eilam has spent the past 24 years at IBM, where she is currently serving as the chief scientist for sustainable computing. Her work involves helping teams reduce the energy consumption of their computing processes. One of her most notable achievements is the open-source project Kepler, which enables the quantification of the energy consumption of individual, containerized applications.In an industry where energy consumption is becoming a critical issue as the AI revolution progresses, Eilam has been ahead of the curve. A Goldman Sachs report this year highlighted that one ChatGPT search requires 10 times the electricity compared to Google Search. Additionally, AI is expected to increase data center power demand by 160% in the near future. Eilam is working with IBM to address these challenges and find sustainable solutions."There needs to be a focus on sustainability in general," she told TechCrunch. "We have an issue, and we also have an opportunity."

The Energy Dilemma and AI's Potential

Eilam believes the industry is facing a conundrum. While AI has the potential to make industries more sustainable, currently, the technology itself is a significant resource drain. However, computing and AI can actually help decarbonize the electrical grid. Currently, the grid relies partly on renewable energy sources like water, the sun, and wind, which fluctuate in price and availability. This makes it difficult for data centers powered by these sources to provide consistent service to consumers."But natural resources aren't her only worry," she said. "Think about how many chips we're manufacturing and the carbon costs and toxic materials involved in the manufacturing process."At IBM, Eilam keeps these problems in mind and approaches sustainable AI holistically. For instance, IBM uses a program sponsored by the National Science Foundation to identify toxic materials in AI chips and accelerate the discovery of new materials to replace them.When it comes to operations, she advises teams on training AI models in energy-efficient ways. "Using less data, but high-quality data, you can converge quicker to a more accurate solution," she said.For fine-tuning, IBM has a speculative decoding technique to enhance inference efficiency. "Then you go down the stack," she continued. "We have our own platform, so we're building a lot of optimizations related to how you deploy these models on accelerators."Eilam emphasizes IBM's belief in openness and heterogeneity. "This is why we released Granite in multiple different sizes. Based on your use case, you can choose the size that suits you best, potentially saving costs and energy."They also build in observability to quantify various aspects, including energy consumption, latency, and throughput. Eilam sees her work as increasingly important as she hopes more people will trust IBM's models for effective and sustainable computing. "What we're telling them is 'hey, don't start from scratch.' Take Granite and fine-tune it. Do you know how much energy you save by not starting from scratch?" she said."The reason they want to start from scratch developing their own models is because they don't trust what's out there. Because you don't know what data went into the training and maybe you're violating some IP," she said. "We have IP indemnity for all our models because we can tell you exactly the data that went in, and we will assure you that there is no IP violation. So, that's where we're saying 'Hey, you can trust our models.'"

A Woman in AI: Overcoming Unconscious Biases

Eilam's background is in distributed cloud computing. In 2019, she attended a software conference where a keynote was about climate change. Since then, she has been committed to merging climate and computing and making a difference. However, diving deeper into AI often meant she was the only woman in the room. She has learned a lot about unconscious biases, which both men and women have in different ways."I think a lot about creating awareness," she said, especially as a woman in a leadership role. She co-led a workshop in IBM research a few years ago, discussing these biases with women. She advises women on their professional journeys to never be afraid to have opinions and express them."Persist, persist. If they don't listen, state it another time, and another time. That's the best advice I can give."

Investor Insights for the AI Future

Eilam thinks investors should look for startups that are transparent about their innovations. "Are they disclosing their data sources?" she said, adding that this also applies to a company sharing how much energy its AI consumes. She also emphasizes the importance of startups having guardrails in place to prevent high-risk scenarios.She is in favor of more regulations, although it can be challenging due to the complexity of the technology. The first step, she says, is transparency - being able to explain what is happening and being honest about the impact."If explainability is not there, and then we're using [AI] without consequences to people's potential future, there is an issue here," she said.