Last week I attended The Artificial Intelligence Conference in NY sponsored by O'Reilly and Intel. It's my yearly ritual to report on the latest academic research, industry trends, successes, and challenges as more organizations invest people, time, and money in machine learning.
From last year's conference I reported here on this blog on my keynote takeaways, and on collaborative AI practices and principles. On CIO, I shared how to get started with deep learning and that deep learning is becoming more accessible to mainstream organizations.
This year, the focus of the conference can be summarized as AI at scale. I saw talks by Twitter, Facebook, Cloudera, and Salesforce on different aspects of scaling ML organizations, practices, and infrastructure. ML cannot scale just by allowing data scientists to select their platform of choice, turn on infrastructure, and run experiments.
Organizations serious about artificial intelligence and machine learning are defining a MLLC - a machine language life cycle similar to an SDLC. They are considering tools, practices, architecture, integration, and governance so that these programs can drive model improvements and research new capabilities.
For now, please have a quick review of my tweet stream below. I'll have more detailed articles coming in CIO and InfoWorld soon.
From last year's conference I reported here on this blog on my keynote takeaways, and on collaborative AI practices and principles. On CIO, I shared how to get started with deep learning and that deep learning is becoming more accessible to mainstream organizations.
This year, the focus of the conference can be summarized as AI at scale. I saw talks by Twitter, Facebook, Cloudera, and Salesforce on different aspects of scaling ML organizations, practices, and infrastructure. ML cannot scale just by allowing data scientists to select their platform of choice, turn on infrastructure, and run experiments.
Organizations serious about artificial intelligence and machine learning are defining a MLLC - a machine language life cycle similar to an SDLC. They are considering tools, practices, architecture, integration, and governance so that these programs can drive model improvements and research new capabilities.
For now, please have a quick review of my tweet stream below. I'll have more detailed articles coming in CIO and InfoWorld soon.
No comments:
Post a Comment
Comments on this blog are moderated and we do not accept comments that have links to other websites.