Today's Tech Digest - Nov 23, 2019

Cheap 5G phones won't come to the masses until these things happen first

One reason why 5G phones cost so much is that the chips cost more, too. Without a 5G-ready chip that can talk to the carrier network, your phone can never reach those lightning speeds. Right now, these 5G chips are tailor-made to each carrier's particular wireless spectrum. So even if you buy the Galaxy S10 5G for AT&T, 5G data won't necessarily work on T-Mobile, Verizon or Sprint. Making 5G phones more or less bespoke to each carrier requires extra time and expense to develop, test and deploy. ... Separate 5G chipsets and modems may not be the norm for long. Qualcomm is working on a way to integrate the two into a single unit. The world's largest mobile chipmaker also plans to eventually make 5G available on multiple carrier bands. Both these changes will simplify what it takes to build a 5G phone, which in turn should make them cheaper to make and maintain. Competition will also help lower the price, especially if players like MediaTek, known for undercutting Qualcomm on processors and modems, can target the 5G midrange chipset market abroad. Qualcomm itself is also committed to making a midrange 5G processor for cheaper phones.

5G: A transformation in progress

The road to 5G began back in 2015, with the ITU's IMT-2020 framework, which set out the general requirements and future development of the next-generation mobile technology (IMT stands for International Mobile Telecommunications) ... The ITU's broad goal for IMT-2020/5G was to accommodate "new demands, such as more traffic volume, many more devices with diverse service requirements, better quality of user experience (QoE) and better affordability by further reducing costs". The key driver for this effort was the need to "support emerging new use cases, including applications requiring very high data rate communications, a large number of connected devices, and ultra-low latency and high reliability applications" ... According to the GSA's latest (January 2019) figures, eleven operators claim to have launched 5G services (either mobile or FWA): AT&T (USA), Elisa (Finland and Estonia), Etisalat (UAE), Fastweb (Italy), LG Uplus (South Korea), KT (South Korea), Ooredoo (Qatar), SK Telecom (South Korea), TIM (Italy), Verizon (USA), and Vodacom (Lesotho). 

Target Sues Insurer Over 2013 Data Breach Costs

In its lawsuit, Target argues that its general liability policy with ACE covers property damage that includes "loss of tangible property that is not physically injured." This, according to Target's lawsuit, includes the replacement of those payment cards because they were "damaged" by the 2013 and could no longer be used. "ACE has refused to acknowledge coverage for the payment card claims and has further disregarded its contractual obligation to indemnify Target for the settlement payments relating to the payment card claims," according to the lawsuit. "ACE has improperly refused to indemnify Target for settlement payments falling within its aggregate coverage layer." ... A Target spokesperson told Information Security Media Group that the company had been negotiating with ACE for a year over this issue before deciding to file the lawsuit in federal court earlier this month. "We believe the costs are covered within the scope of the insurance policy Target has with ACE and are focused on resolving the outstanding claim," the Target spokesperson says.

Extreme targets data center automation with software, switches

Extreme Fabric Automation is hosted as an application on a guest virtual machine of the two new switches, providing on-premises and private-cloud deployment options, said Dan DeBacker, director of product management at Extreme. “The idea is to remove the need for IT to have to do manual switch-by-switch configurations,” he said. In addition, the software gives IT teams the ability to scale the network up and down to meet changes in demand, and it reduces the cost of operating the network. For those using the guest VM, it eliminates the need for an external server, DeBacker said. The Extreme Fabric Automation package now integrates with orchestration software including OpenStack, VMware vCenter, and Microsoft System Center Virtual Machine Manager (SCVMM). Each integration is a separate microservice and additional integrations will be available in future releases of the software, Extreme said. The orchestration software further automates network configuration, coordination, and management of resources, DeBacker said.

Why SaaS-based AI training will be a game changer

What strikes me about this approach to AI training is that you need a sound training data set. In some cases, it can be obtained from open or proprietary training data brokers. In most instances, you format your own data to train the machine learning model. However, what if other trained machine learning models could train models, anywhere and any time? The idea is not new. Since the advent of AI we’ve toyed with the idea of having one AI engine teach another, either by sharing training data or, better yet, sharing knowledge and experience through direct, automatic interaction. Having one AI engine mentor yours provides outside experience and thus makes the AI model more valuable and effective. This is easier said than done. Machine learning engines typically don’t talk to each other, even if they are the same software. They are designed from the ground up to be stand-alone learners and interact with non-AI systems or humans. However, inter-AI engine training is on most vendor radar screens.

Read more here ...