Facebook VP: AI has a compute dependency problem

Facebook VP: AI has a compute dependency problem

At Transform 2019, Facebook VP Jérôme Pesenti said more software optimization will be required for artificial intelligence to keep growing.

On Friday July 12, 2019, Khari Johnson at VentureBeat wrote:
“The peak complacee for companies like Facebook and Google can afford for an experiment, we are reaching that already.”

Analysis introduced last month found that the costs of training systems like OpenAI’s GPT-2 can exceed carbon emissions of the lifetime of five cars.

Many Facebook services are powered by Intel CPUs, Facebook engineering manager Kim Hazelwood said last year.

“My goal is to make every single engineer in the organization an ML engineer, and that number has increased 3x in the last year, so you’re talking about thousands and thousands of engineers that are not on my team and are actually not actually ML engineers,” he said.

About the Author:

Khari writes about artificial intelligence for VentureBeat and lives in San Francisco. Nachos make me happy. I adhere to VentureBeat’s ethics policy. This is a complete list of VentureBeat articles written by Khari Johnson, in reverse chronological order. Khari Johnson has recently written 5 articles on similar topics including machine learning, science:
  • "TensorFlow Lite 1.0 for lightweight machine learning on mobile and IoT devices made its debut today with a number of improvements and shared a dev roadmap". (March 7, 2019)
  • "Google introduced the latest version of TensorFlow today with an emphasis on ease of use and a reduction in APIs for the machine learning framework". (March 6, 2019)
  • "Google Cloud today announced that Andrew Moore, dean of the Carnegie Mellon University School of Computer Science, will lead its AI efforts in the future. Moore takes over the role of Google Cloud’s AI chief scientist from Dr. Fei-Fei Li, a Stanford University professor who has occupied the role since 2016. Moore, who announced plans to …". (September 10, 2018)
  • "If Google's conversational AI Duplex is a harbinger of things to come, bot-to-bot communication will soon be critical, which is why the makers of assistants like Siri, Alexa, and Google Assistant need to work together". (July 6, 2018)
  • "Google should give away Home Mini smart speakers to fend off Amazon Echo and maintain its dominance in search, a Morgan Stanley analyst argues". (June 29, 2018)