Re: MLCommons Leadership Evolution
David Kanter
Together we built the industry standard for measuring AI performance and energy efficiency and helped drive the industry to deliver >50X performance gains since our inception, a pace that is astounding compared to Moore’s Law. What started with MLPerf Training has blossomed into a rich portfolio that spans from the smallest microwatt-scale IoT devices to the smartphones we all carry up to multi-megawatt supercomputers. We’ve also developed a dynamic and growing community in areas ranging from datasets to medical AI.
I am excited about the opportunity to focus on helping MLPerf and our working groups reach the next level of impact. For MLPerf Training and Inference we are staying abreast of the ever-changing workloads and use cases driven by the evolution of AI while increasing customer adoption with tools such as Collective Minds and MLCube.
Today we also see the rapid shift towards running AI on client devices, and our team on MLPerf Client is at the forefront there. My vision is to have client benchmarks that every MLCommons member can run on their favorite devices, from cell phones to PCs. Similarly, the automotive community is building vehicles that are increasingly intelligent and autonomous and MLPerf Automotive can help guide the community towards safer and more efficient transportation. We also have tremendous opportunities in networking, storage, and inference APIs.
I am excited about continuing my work with all of you to continue to build on the success of MLPerf and MLCommons. As always, feel free to reach out - I’ll look forward to hearing from you.
Thanks,
Hello everyone,I’m honored to step up as the interim Executive Director of MLCommons. As Peter mentioned, I’ve been working for MLCommons for the past 6 months as a Distinguished Fellow on the AI Safety effort. I have been incredibly impressed by the organization, its members, and the dedicated commitment from all of you on the important work we are doing. Core to MLCommons is the MLPerf benchmark work we do and having David as a dedicated partner to drive that work forward will ensure MLPerf continues to grow and remain the pre-eminent performance benchmark for AI Systems.
In line with our mission of building Better AI for Everyone, MLCommons has a number of emerging efforts that build upon MLPerf’s success. It is the strength of MLPerf that allows us to extend into these these exciting new areas such as AI Safety, medical benchmarks, and our Data efforts. I’ll continue to deepen focus in these new areas as we move towards delivering a v1.0 AI Safety benchmark suite later this year, while also driving growth and communication strategies for these other newer workstreams.
I believe MLCommons has a critical role to play as AI measurement science matures. I spent over 10 years driving real world impact at Mozilla, building teams from the bottom up and developing data science and engineering strategies supporting cross functional initiatives across product and marketing teams, as well as revenue and growth programs. The impacts of AI will be felt across the economy and through all of society, and it is only through better measurement that we will be able to drive improvements in the technology and its deployment. But advances in measurement science alone will not be enough, and I believe MLCommons is the organization best-prepared to operationalize and standardize cutting-edge research to drive real-world impact across the industry.
With the rapid advancement in AI, our work is more important than ever. I look forward to working with all of you, and my virtual door is always open. Please feel free to reach out to say hi!Best,RebeccaOn Wed, Jun 26, 2024 at 1:52 PM Peter Mattson <peterm...@google.com> wrote:Dear MLCommons Community,
The MLCommons Association is evolving its leadership structure to reflect our growing role in the AI ecosystem, and to enable future impact.
Thanks to the hard work of our dedicated community, the scope and impact of our organization is growing.
Since the early days of MLPerf, our community has taken on bigger and bigger challenges. We’ve grown from one performance benchmark, to a family of performance benchmarks, to an organization with a mission to make “AI better for everyone.” Beyond our core MLPerf work, our efforts have expanded significantly - ranging from a dataset format standard now in use by multiple major portals, to an innovative federated evaluation platform for benchmarking medical AI accuracy now in discussion with leading pharma organizations. And most recently, we have initiated an effort to develop a global standard for AI safety benchmarking with dedicated support from industry sponsors. The AI Safety effort is of significant interest to policymakers and standards bodies worldwide, and is on a fast trajectory to deliver a production level benchmark suite later this year. We believe this effort has the potential to substantially benefit society and the AI industry broadly by increasing AI safety and reducing purchasing and compliance uncertainty.
To ensure all efforts are fully supported for maximum success, we are evolving our organizational leadership. We have a strong leadership team to lead us into this next phase of growth.
Rebecca Weiss will be stepping up as interim MLCommons Executive Director. She is already deeply familiar with MLCommons in her role as a Distinguished Fellow as part of the AI Safety leadership team. Rebecca brings a wealth of experience including starting and building the data science function at Mozilla where she also conceived and built Rally. She currently serves on the Board of the Pew Research Center, and holds graduate degrees from MIT and Stanford.
David Kanter, who helped build MLCommons from the ground up during our transformative early years, will take on the role of Head of MLPerf, reporting to the BoD. David is a widely-recognized leader, spokesperson, and technical expert for our industry standard performance benchmarking. MLPerf is a core priority for the organization and with this dedicated focus from David we will be able to deepen support as we continue to grow.
If you have any questions about this change, please don’t hesitate to contact the board at bo...@mlcommons.org or Rebecca, David, or myself.
MLCommons’ efforts would not be possible without the support of our members and community. Thank you all.
Kind regards,
Peter
