ArchiTECHt Daily: Intel is doubling down on computer vision

So there you have it: "Intel Inside" might very well become "Intel Under the Hood" over the next deca
ARCHITECHT
ArchiTECHt Daily: Intel is doubling down on computer vision
By ARCHITECHT • Issue #36
So there you have it: “Intel Inside” might very well become “Intel Under the Hood” over the next decade or two, as the chip giant announced on Monday that it’s acquiring smart-car company MobileEye
Notably, MobileEye focuses on computer vision systems for cars. This follows in the footsteps of Intel’s September acquisition of Movidius, a startup that builds specialized chips and algorithms for computer vision. 
Intel missed mobile and might largely be playing catch-up with Nvidia in the world of artificial intelligence right now, but as I wrote just a couple weeks ago, there’s a big opportunity for Intel to own applied AI workloads. Our devices, wearables, cars and other consumer goods are going to need intelligent computing locally, not just in the cloud, and it seems unlikely they’ll all get it via GPU. Combine all the research being done to develop specialized AI processors with Intel’s deep pocketbook, and you would seem to have a recipe for success. 
Of course, Intel still needs to execute (and spend wisely), and AI still needs to go mainstream, but those are discussions for another day.
P.S. Stratechery has a really good business-case analysis for this deal here.

Sponsor: Datos IO
Around the web: Artificial intelligence
I am going to try this out today with a podcast interview. I’ll report back about how good, or painful, the experience was.
Cuban might ultimately be right, but for now I’m filing this under “hyperbole about AI coming out of SXSW.” Right alongside this theory on AI being a “fascist’s dream.” (Although that seems less hyperbolic today than it would have a year ago.)
Marketing copy from Microsoft, but a good example of low-hanging fruit in utilizing AI services to classify research reports and. Incorporating humans, rather than automating the whole process, is probably what makes it work.
This research from OpenAI is just the latest to adopt evolutionary algorithms for training reinforcement learning models, to automate the process.
arxiv.org  •  Share
I’ll listen anytime Michael Jordan and Ion Stoica have their names on a paper. Given everything that’s come out of Berkeley, including Spark, it’s fair to assume they could advance machine learning yet again.
arxiv.org  •  Share
Around the web: Cloud and infrastructure
I’m pretty sure Google dinged AWS last week about not offering this, so here you are.
To be clear, the work was done by Google engineers not Netflix, but bringing these tools 
Sponsor: Marshal.io
Around the web: All things data
I think the data science stuff will be less impactful overall than the cloud services. Cloudera, like its peers, needs a play in the cloud.
Technology aside, this is a great use case. We cannot apply enough technology to this cause, as far as I’m concerned.
Did you enjoy this issue?
ARCHITECHT
The most interesting news, analysis, blog posts and research in cloud computing, artificial intelligence and software engineering. Delivered daily to your inbox. Curated by Derrick Harris. Check out the Architecht site at https://architecht.io
Carefully curated by ARCHITECHT with Revue. If you were forwarded this newsletter and you like it, you can subscribe here. If you don't want these updates anymore, please unsubscribe here.