60,000 patents > ​@Microsoft open-sources its patent portfolio https://t.co/rLsfKcF6Zp

Notice that the official announcement was published on the Azure blog: https://t.co/630IthRMnE

Now @google on stage at @TheAIConf suggests to use “ML Ops” to mean “DevOps for ML”.

Just in case “DevOps” alone was not problematic enough.

Interesting closing remarks from @IBMWatson at @TheAIConf:

“Explanation is the new UI for #AI”

@AklWaleed Not according to @intel, who keynoted the event and spent a lot of effort in promoting their new OpenVINO toolkit (soon open source) to process ML workloads across CPUs and GPUs to enhance performances.

On stage at @theAIconf, @IBMWatson describing its new bias detection tool says “you want to balance between fairness and accuracy”.

I don’t think that the notion of “more fairness, less accuracy” is going to fly with non-tech people (especially who is not aware of human bias).

The strategy, product design, marketing and sales for enterprises is completely different from end user market (especially for gaming). It’s not just a matter of adapting #AI tech to different use cases.

Danny Lange, VP of #AI & ML at @unity3d says on stage at @TheAIConf that Unity is a gaming company but mission is much broader than that.

All gaming platforms have the potential to leverage their tech for research, military and enterprise app. Problem is that it’s not their DNA.

Another reason with #AI investments in the enterprise is still very limited: they require an army of experts.

Courtesy of @microsoft at @TheAIConf https://t.co/326heOfAUx

Focus mismatch at @TheAIConf: the keynote speaker is talking about changing the world (before promoting an accounting tool) while an attendee is browsing a website about CPU benchmarks.

Investments in #AI in the enterprise is still minor. Maybe because most tools are impossibly complex and require skills close to black magic? #TheAIConf https://t.co/bm6RhqK38T