다음을 통해 공유

How We Share the Latest AI & ML Developments Within Microsoft

  • 아티클

We recently concluded the Fall 2017 Machine Learning, AI & Data Science (MLADS) conference, Microsoft's largest internal gathering of employees focused specifically on these areas. This latest edition was the eighth in a popular series that we launched back in 2014. Over 3,500 employees tuned into the sold-out conference, both in person in Redmond and over livestream throughout the world, and thousands more will catch recordings of MLADS sessions over coming months.

As the use of AI and ML explodes both within Microsoft and in our customer-facing products and services, our community interest groups catering to these areas have also mirrored that rapid expansion. The MLADS conference itself is unique in that it is almost entirely driven by enthusiastic community volunteers – a band of employees unified in its passion for AI and ML, and a desire to network and learn from one another. The "call for content" that goes out for this conference series routinely gets several hundreds of submissions, and our volunteer team helps triage these submissions and curate the best ones for our event.

The fall 2017 conference featured over 95 talk sessions, 20 tutorials and 65 poster/demo sessions covering a gamut of topics including Azure Machine Learning, advanced Deep Learning, Generative Adversarial Networks, Natural Language Processing, Text Mining, Vision, Speech, Cognitive Services, Time Series Methods, Forecasting, Azure Batch AI, Design & Interpretability, the very latest AI tools, and numerous applications of AI – now even in the generation of original music and creation of original art.

Poster/Demo Reception at the Fall 2017 Microsoft MLADS conference

Talks and tutorials about open source offerings from Microsoft and third parties are an integral part of the conference, and this time was no exception.

A panel of volunteer judges help us shortlist top submissions for our Distinguished Contribution Awards, an internal recognition program for top-class work being done in this field. Selective submissions are also chosen for inclusion in the prestigious [internal] Microsoft Journal of Applied Research.

The conference opened to packed keynote sessions by Joseph Sirosh, Corporate Vice President of Cloud AI, and Todd Holmdahl, Corporate Vice President of Quantum Computing. Joseph's keynote, on the AI Tidal Wave, highlighted the impact AI can have (and is having already) on the lives of people all around the world. His keynote featured leading-edge demos showing some of the latest AI and ML-led innovations happening across Microsoft.

Joseph Sirosh, Corporate Vice President of Cloud AI, Microsoft

As part of Todd's keynote, about Empowering the Quantum Revolution,we pre-announced the Microsoft Quantum Development Kit (QDK), a set of development tools for quantum developers. These tools were subsequently made available to the public, visit this web page for more information.

Todd Holmdahl, Corporate Vice President of Quantum Computing, Microsoft

The Fall 2017 MLADS conference had several other notable features, including:

  • Introductory talks and tutorials to get newcomers caught up on these areas. This set of sessions and a couple of related programs help developers who are not formally trained in AI and ML quickly ramp up and get their hands dirty through real projects.
  • A set of immensely popular sessions from colleagues who have joined our community relatively recently, including from LinkedIn (based in Silicon Valley) and Maluuba (based in Montreal).
  • A Startup Showcase featuring the latest set of exciting startups from the Microsoft Accelerator program in Seattle.
  • A Robotics Expo, featuring many new ideas and experiments involving the latest in ML and robotics.
  • Informal lunchtime meetups on a variety of topics, including ML for sports predictions, ethics and AI, AI in neuroscience, and AI for Earth.
  • A closing plenary session on Intelligent Design / Design for Intelligence, by Bill Buxton.

The fall event also featured our first Family Build-a-Robot Night – children from many Microsoft families came together to build and program wheeled robots, which they then then tested on a variety of tracks. This was just as much fun for our teachers and volunteers as the kids themselves!

Between bi-annual MLADS conferences, we also drive monthly community events and periodic hackathons. These investments are helping our fast-growing (and global) group of AI and ML practitioners learn from each another, even as we continue to rapidly enhance Microsoft's products through the power of everyday AI.

ML Blog Team