Recap of the 2020 CMG IMPACT Conference

I spoke at the CMG IMPACT conference. My talk was about recent advances in machine learning and AI that can be useful for IT organizations. The event ran February 9-13 in Las Vegas. See https://cmgimpact.com/.

The IMPACT conference is put on by an organization called CMG which stands for Computer Measurement Group. See https://www.cmg.org/tag/metrics/. The group’s Web site says that CMG is a not-for-profit organization composed of thousands of technical experts. The group has been around for over 40 years — longevity like that is quite impressive.



This was the first time I’ve attended and spoken at CMG IMPACT. Let me summarize and state that I really liked the event and it exceeded my expectations. The talks were excellent, the vendors had interesting products, I learned a lot while talking to attendees, and the event was very well organized. My compliments to event organizer Amanda Hendley, and logistics person Michelle Cervantes.

I estimate there were about 125 attendees, speakers, and vendors at the event. The event expo was small (about a dozen companies) but they were all interesting and had knowledgeable representatives.

The conference attendees had very diverse backgrounds. There were people from large companies such as IBM and Netflix, and there were people from mid-sized and small companies too. There were people who had highly technical roles, such as performance analysis engineers, and people in semi-technical roles such as business development. The mixture of backgrounds worked quite well.



My talk was titled “Four Recent Advances in AI and Machine Learning”. I thought my talk went well and it received one of the “best-of” awards as voted by attendees — hooray me!

There were many good talks at the event. I especially liked “Among a Billion Configurations, How Would You Find the Best One?” by Stefano Doni, and “Netflix Performance Tales in One Take” by Ed Hunter, and “Deep Learning’s Most Dangerous Vulnerability: Adversarial Attacks” by Luba Gloukhova. But literally all the talks I listened to at CMG IMPACT were very good in my opinion.

At the beginning of my talk I asked the audience if there were any neural network skeptics in the room and several people raised their hands. Then at the end of my talk I asked one of the skeptics, a gentleman named Igor, to pose me a difficult question. He suggested that neural networks don’t really “understand”.

I responded that Igor was spot on in his observation and that there’s a growing sense that traditional neural architectures are reaching the limits of their capabilities. And, viola! I had a slide ready that showed how there’s increasing interest in neuromorphic computing, which more closely resembles biological systems and which, in theory, might lead to systems that “understand” in some sense.

Anyway, the 2020 CMG IMPACT conference was a good use of my time. I’m confident I represented my company well and I picked up information that’s directly applicable to my job. I will definitely try to attend and speak at next year’s event.

This entry was posted in Conferences. Bookmark the permalink.

1 Response to Recap of the 2020 CMG IMPACT Conference

  1. Thorsten Kleppe says:

    Congratulations for the award you won Dr. McCaffrey.
    How I would like to have listened to your lecture. So far, your talks that were freely available on the Internet are always very rich in valuable information.

    The question from Igor is really interesting and seems to me related to the topic of the “Hermeneutic circle”, there could be a deep relationship to ML and to us.
    It’s really crazy to me if I think about what was there first for my understanding, the letters, the words or the sentences. My feeling would bet on the words, my logic would bet on the single chars.

    But my understanding seems to be a illusion, because many topics feels today completely different as at the first time. How can I be sure that I understand, or is, what we think what understanding is, just a approximation like a neural Net? A neural Network need this more cycles too for more generalization.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s